Daniel T. Halperin is an epidemiologist and adjunct professor at the University of North Carolina’s Gillings School of Global Public Health. He co-authored “Tinderbox: How the West Sparked the AIDS Epidemic and How the World Can Finally Overcome It.”

As lockdown restrictions ease, a critical question looms: When do we reopen schools? Parents and others weighing covid-19′s risk to children and the adults they may infect, directly or indirectly, should consider emerging evidence that suggests children are not significant transmitters of covid-19. These data, coupled with the enormous adverse impacts of continuing closures, argue for reopening schools by fall.

Of about 360,000 covid-19 deaths worldwide, only about two dozen children are known to have died. For all the recent reports of serious complications among young people, these are statistically rare and, if detected early, most afflicted youths recover within weeks.

While most countries have shuttered schools, others such as Taiwan have achieved effective responses without closures. In Denmark and Norway, where schools began reopening in mid-April, covid-19 cases and deaths have decreased. Normally, gregarious youngsters are efficient spreaders of respiratory pathogens. But this appears not to be the case with covid-19.

Emerging evidence suggests that, much like with the Severe Acute Respiratory Syndrome (SARS) epidemic in 2003, children are less likely to become infected with this coronavirus. From Feb. 12 to April 2, just 1.7 percent of U.S. cases for which age is known occurred among people younger than 18. Some researchers theorize that some resistance has been conferred by previous exposure to other coronaviruses, such as those that produce the common colds that children frequently acquire. Additionally, a study published in JAMA found that youths are less prone to infection because they produce smaller quantities of a protein, ACE2, which both SARS and the novel coronavirus use to enter cells.

A German study that warns against reopening schools found viral loads in infected children at levels comparable to adults. There is evidence, however, that as with the earlier SARS outbreak, children who have covid-19 are less contagious than adults. Many children with covid-19 are asymptomatic; in the absence of coughing and sneezing, they emit fewer infectious droplets. Remarkably, contact tracing studies in China, Iceland, Britain and the Netherlands failed to locate a single case of child-to-adult infection out of thousands of transmission events analyzed. A review of studies from several Asian countries identified few cases of children bringing the virus home, and a recent analysis of covid-19 interventions found no evidence that school closures had helped contain the epidemic.

Some of this data likely underestimates children’s potential to infect others because information was collected after lockdowns and other mitigation measures were implemented. Still, the findings from contact tracing and the significant biological differences between covid-19 and more common respiratory ailments suggest that children are not major sources of infection.

Anthony S. Fauci, director of the National Institute of Allergy and Infectious Diseases, recently warned against reopening schools too early, and noted complications in some children that resemble Kawasaki disease. The emerging condition, known as multisystem inflammatory syndrome in children, is troubling and must be monitored. But it also appears rare; so far, only several hundred U.S. cases have been reported.

The low numbers of children affected by covid-19 and the new syndrome should be considered in additional context: More than 200 U.S. children were killed last year by flu; some 10,000 others died from various childhood diseases. A rare condition that is not commonly fatal does not justify keeping 55 million American students home into the next academic year.

In March, Imperial College of London modeling estimated that 2 to 4 percent of covid-19 deaths in Britain might be prevented by closing schools and colleges, compared with a potential 17 percent to 21 percent prevented from self-isolating. This suggests that schools are not particularly significant contributors to community transmission. As schools reopen, one concern is for the risk to groups already known to be vulnerable to covid-19: predominantly older people with predisposing conditions such as chronic illness, obesity and smoking. The options are distressing: potentially greater numbers infected vs. the rising educational costs to millions of children.

School closures of course affect more than academics. Students are also deprived of social connections and physical activity. Socioeconomic disparities are exacerbated, as some families have resources to enhance online learning, while less privileged children fall further behind.

Other consequences of school closures include recent surges in child abuse; hunger from missed subsidized meals; greater anxiety, depression and isolation. Students with autism, Down syndrome, attention-deficit hyperactivity disorder and other special needs are at particular risk. But months away from friends and school structure takes a toll on all students, as beleaguered parents everywhere can attest.

Schools have begun reopening in France, Germany, Switzerland, Japan, Australia and elsewhere. Adequate testing and evidence-based safety precautions are essential for protecting teachers and other staff. Although some covid-19 cases regrettably may result from reopening schools, the existing evidence does not warrant inflicting potentially long-term academic, social and vocational disadvantages on millions of children.

Read more: