Jean Gaumy / Magnum

Given that bars and restaurants in many parts of the United States are beginning to reopen, while the prospects for school remain hazy almost everywhere, you might think that scientific evidence about kids and the coronavirus is nonexistent. The truth is that we are still somewhat in the dark, but not completely. Here’s what we know, what we kind of know, and what we need to do to know more.

Back in February and March, when the pandemic was in its early stages, the big question was whether kids were at high risk for COVID-19. It wouldn’t have been surprising if they were; other respiratory diseases such as the flu are known to significantly affect both children and the elderly. But one of the robust findings about COVID-19 in the past few months is that children are among the least affected groups. They are less likely to contract the disease, and if they do contract it, they are more likely to have a mild or asymptomatic case. Death rates are much lower. This evidence doesn’t mean that kids cannot get sick, and or cannot fall seriously ill, but older adults are far more susceptible.

The other big question was: Are children major vectors for the virus? At least one government has argued that children could not transmit the virus at all. Some research teams countered that they were just as likely to transmit it as adults are. At this point, neither of these claims seems quite right. If a kid is sick and shedding virus particles and an adult is exposed to those particles, of course the adult can get sick; but children do seem to transmit the virus less than adults do. In an early case, an infected child went to several skiing schools and was exposed to hundreds of contacts without infecting anyone. Data from the Netherlands suggest that children are relatively unlikely to be the “index case” in their families—that is, they are unlikely to be the first case in a family cluster.

If kids are probably low-risk (in terms of both getting sick and transmitting), that doesn’t necessarily mean states should go ahead and reopen schools. That’s because schools do not contain only children. This is not Lord of the Flies. The adults at schools may be at risk from interacting with kids, but also from interacting with one another, and with parents, and with other adults as they travel to and from work.

We have some information from abroad. France, Germany, Denmark, and other countries have reopened schools. Sweden has had schools open the whole time. Oddly, one of the most compelling pieces of evidence is provided by what we haven’t seen: much in the way of large-scale outbreaks linked to schools. Some cases, yes—but not super-spreader events like the ones documented all over the world at bars and meatpacking plants.

Beyond what we haven’t seen, some early information on adults at school is encouraging. In Denmark, some preliminary data suggest that teachers are not an especially high-risk group. A recent report out of Sweden looks at risks of exposure to COVID-19 by occupational group, and notes that school staff are not more likely than other occupations to contract the disease. Preschool and high-school teachers are actually less likely to get COVID-19. The highest-risk group here is drivers—of taxis and buses in particular.

An exception is Israel, where the school-reopening process has been up and down. Israel opened schools in May, but subsequently closed a number of them temporarily after detecting cases. The country had one large outbreak tied to a school. Perhaps Israel is faring less well than European countries because it opened with fewer social-distancing measures. But even in Israel, the total count of cases tied to schools since they reopened stands at about 300—a very small share of the country’s students, teachers, and staff.

The above does not amount to airtight evidence—I’ve gleaned this information from a close reading of news reports, which is not how data gathering should work. I should not be trying to answer the question “What is going on in schools that reopened?” by Googling around; I resorted to that method because of the absence of a publicly available data set derived from a universal school-based testing regime. Some countries are collecting good data: In Germany, at least some schools are testing kids and teachers twice a week. This is great, but whatever Germany has found, it hasn’t yet shared with the public.

If countries with open schools simply reported the number of confirmed COVID-19 cases per school each week—if any—that would do wonders. In the U.S., schools are closed but some child-care centers and camps are open. Local governments should be collecting data from these sources. I started doing this—in an unscientific and nonrandom way—simply out of frustration that no one else was. This lack of information-gathering perhaps shouldn’t be surprising, as the overall pandemic response in the U.S. has been worse than elsewhere; we have fallen down on testing, contact tracing, and everything else.

Even minimal data could start to answer certain lingering questions. Does age group matter in transmission rates? (One difference between Israel and countries such as France and Sweden is that Israel opened all its schools at once, and others started with younger children.) Which prevention measures matter? Do kids need to wear masks and socially distance, or is conscientious hand-washing and mask wearing for teachers enough?

The fact is, parents can’t wait around forever. As long as they have to stay home with children, they cannot truly participate in the workforce. The facts right now suggest that reopening schools would not lead to disaster, but more information shouldn’t be so hard to come by.

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