On Wednesday, New York City — with the largest public school district in the country — suspended in-person learning because of rising covid-19 cases in the community. Mayor Bill de Blasio (D) had said that schools would close if the positivity rate of coronavirus tests in the city passed 3 percent, and it did.

But while rising test positivity is a cause for concern and a reason to increase public health interventions, it can’t be looked at in isolation — and it’s not a good measure on which to base a decision to close schools. The metric is best interpreted as a gauge for how hard we’re looking for cases and how hard it is to find them. It is not equivalent to prevalence of the coronavirus in the community (although it is related to that figure); mainly, it’s a signal to do more testing.

The only epidemiological reason to close schools would be evidence that they are magnifying infection and driving community spread. And so far, there is no data suggesting that in-person education is doing those things. To the contrary, the prevalence of infection in schools is comparatively low.

Given the long-term harms that occur when children are out of school for extended periods (from lower lifetime income to shorter life spans), keeping schools open, as Europe is doing even during its new wave of lockdowns, should be the default. It’s become a cliche to note that bars, restaurants and gyms should close before schools do, but that doesn’t detract from the truth of the observation. De Blasio said this week that it is “just a matter of time” before the city asks restaurants to close down, but that is cold comfort to parents who must scramble to fulfill their children’s educational needs.

Obviously, this argument would not hold if the virus were rampant in schools or streaming from schools to the community. But it isn’t. To date the data from New York City schools reveals a generally optimistic picture. In the first week of its testing program, begun last month, only 28 tests out of more than 16,000 conducted returned positive.

Nor is there anything inherently special about a 3 percent positivity rate. It’s an arbitrary metric — and, what’s more, it’s based on testing throughout the city, not schools in particular. It’s also on the conservative side: The World Health Organization has said that governments can consider opening up some activities when the rate is below 5 percent. The positivity rate within New York schools, where an aggressive testing program was in place, was below 0.3 percent.

A high positivity rate can be a sign that a city or state is testing too few people: only those who have presented at a doctor’s office or hospital with symptoms, for instance. But that does not appear to be the case in New York state, which has among the highest rates of testing in the country: 8.2 tests per 1,000 people, according to data from the Covid Tracking Project. In New York City, virtually all Zip codes have adequate levels of testing. At reasonable levels of testing, a rising positivity rate does generally mean that community spread is rising, but schools had largely been spared.

To be sure, remote learning can be effective in some instances, but not every student excels at it. Primary school students, students with learning disabilities, nonnative English speakers and low-income students without access to decent computers and WiFi at home suffer in particular. Many private schools, including religious schools, will remain open — another way the covid-19 shutdowns will magnify the gap between children from lower-income and higher-income groups.

If the pandemic response in this country had not been handled so badly by the federal government and the states, we would not be arguing about schools at this point. But the bungled responses mean that politicians face tough decisions — and questions of prioritization. New York City appears to have prioritized keeping the wrong institutions open. As we combat the virus, we need to think about both short-term and long-term effects. Disruption in children’s learning — at a critical period in their cognitive development — will have dire long-term effects.

In New York, the teachers’ unions have pushed back against opening the schools (and now, keeping them open) on safety grounds. As a product of a public high school who deeply values the interactions I had with so many teachers, I respect their concerns. To me, they are essential workers. Schools do need more money to keep student density down, provide masks to students who lack them, improve ventilation in some older buildings and protect older teachers. But New York schools had in place rigorous plans for social-distancing and staffing, and they seemed to be working. The city can work with teachers and their unions to alleviate their remaining concerns.

New York City, once the nation’s epicenter for the coronavirus, quelled a ferocious outbreak and started to come to life again. It has since faced a setback. Yet while indoor dining and workouts continue, students once again can’t sit in classrooms and discuss literature, math and science with their teachers. Given the lack of evidence that schools were spreading the virus, something is deeply wrong with that picture.