Originally published at NationalReview.com.

Just over one year ago, on March 12, 2020, the children of Ohio were the first in the nation to be sent home from shuttered schools because of COVID-19. Within two weeks, every other state had followed Ohio’s lead. Given how little we knew about the virus then, most people believed closing schools was the safest decision. But a year later the evidence says something entirely different: Closing schools was a mistake.

The high costs of closing schools are indisputable. Surveys of parents and teachers, as well as projections and trends in test scores, indicate that students are far behind in their learning, and that achievement gaps are growing. Economic estimates predict that students will lose $12,000–$15,000 in lifetime earnings for every month schools remain closed, and related U.S. GDP losses are forecast to run into the tens of trillions of dollars.

Even more important costs cannot be measured in dollars or test scores. The social isolation that students have struggled with in this “new normal” has led to widespread increases in child anxiety and depression, punctuated by tragic rises in adolescent mental-health emergencies and suicides. These costs are a strain for most students, and almost inconceivable for the very youngest, who may struggle to remember life without them.

Had blanket school closings been absolutely necessary to control COVID-19, their costs could be justified in retrospect. But we know now that they weren’t absolutely necessary. Children are roughly half as susceptible to COVID-19 as adults, and are far less likely to experience severe symptoms or increase transmission among adults. CDC data indicate that school-aged children make up under 10 percent of confirmed COVID-19 cases and 0.08 percent of COVID-19 deaths, despite accounting for 23 percent of the population. And study after study shows that, when basic mitigation strategies are followed, in-school transmission is exceptionally rare.

Even for those who don’t trust the research, evidence that universal school closures could have been avoided is in plain sight. Data from my recently launched Return to Learn Tracker, which monitors over 8,500 school districts, show that thousands of districts have been fully in-person since November 2020. In fact, many districts stayed open when COVID-19 caseloads were three to six times last spring’s highs. Yet even under the watchful eye of school-district leaders, researchers, and the media, accounts of increased transmission in these districts haven’t materialized.All Our Opinion in Your Inbox

Closing schools last spring should be identified as the mistake it was. It was a reasonable step to take at the time, because of the uncertainty we then had about the lethality, transmission, and best practices for mitigation of the virus. But we must name it a mistake now, because, in too many places, reopening decisions remain rooted in a conviction born last spring that universal closures are the only truly safe option. And until leaders with the gravitas to challenge that conviction do so, too many reopening decisions will remain grounded in the past.

This problem is rooted in what social psychologists call “motivated reasoning,” whereby we give more weight to evidence supporting our views and discount that which challenges our priors. Faulty “motivated reasoning” is amplified by the political polarization that has characterized reactions to the COVID pandemic since last summer, so for many on the left, President Trump’s repeated, ham-fisted demands that schools reopen immediately last summer only cemented the belief that in-person learning is unsafe. Indeed, the partisan divide over school reopening in opinion polls is just as stark in districts’ actual offerings. Four times as many districts in counties that voted for Biden (21 percent) as districts in counties that voted for Trump (5 percent) continue to keep all their schools fully remote.

Changing perceptions and decisions in those blue districts will require strong Democratic leadership, particularly from President Biden. But so far, that leadership has been inconsistent.

Biden’s December promise to get a “majority” of schools reopened in his first 100 days was a strong start, but the administration faltered in following through. In February, White House Press Secretary Jen Psaki dialed back the pledge, saying the goal was now to have “more than 50 percent” of schools open for “some teaching in classrooms, at least once a week.” A week later, President Biden said he hoped to be “close” to reopening K–8 schools by April 30, “many of them five days a week.”

Biden must lead with conviction and take advantage of the strong tailwinds right now — falling case rates, a trend toward reopening schools, and mounting evidence on how to do so safely — to challenge outdated beliefs about the risks of in-person learning. Vague hopes of being “close” to reopening won’t get the job done. The president needs to unequivocally set the expectation that all students be given the opportunity to attend school in person by the end of April.

The science and evidence are clear: Closing schools last year was a reasonable reaction to an impossible set of circumstances, but it was a mistake, and that mistake will stay with us until we admit it. In March 2021, we need the leadership to act on what we now know, and to deliver the option for every student to safely resume in-person learning this year.