I am no longer a professor of medicine at Harvard. The Harvard motto is Veritas, Latin for truth. But, as I discovered, truth can get you fired. This is my story—a story of a Harvard biostatistician and infectious-disease epidemiologist, clinging to the truth as the world lost its way during the Covid pandemic.

On March 10, 2020, before any government prompting, Harvard declared that it would “suspend in-person classes and shift to online learning.” Across the country, universities, schools, and state governments followed Harvard’s lead.

Yet it was clear, from early 2020, that the virus would eventually spread across the globe, and that it would be futile to try to suppress it with lockdowns. It was also clear that lockdowns would inflict enormous collateral damage, not only on education but also on public health, including treatment for cancer, cardiovascular disease, and mental health. We will be dealing with the harm done for decades. Our children, the elderly, the middle class, the working class, and the poor around the world—all will suffer.

Schools closed in many other countries, too, but under heavy international criticism, Sweden kept its schools and daycares open for its 1.8 million children, ages one to 15. Why? While anyone can get infected, we have known since early 2020 that more than a thousandfold difference in Covid mortality risk holds between the young and the old. Children faced minuscule risk from Covid, and interrupting their education would disadvantage them for life, especially those whose families could not afford private schools, pod schools, or tutors, or to homeschool.

What were the results during the spring of 2020? With schools open, Sweden had zero Covid deaths in the one-to-15 age group, while teachers had the same mortality as the average of other professions. Based on those facts, summarized in a July 7, 2020, report by the Swedish Public Health Agency, all U.S. schools should have quickly reopened. Not doing so led to “startling evidence on learning loss” in the United States, especially among lower- and middle-class children, an effect not seen in Sweden.

Sweden was the only major Western country that rejected school closures and other lockdowns in favor of concentrating on the elderly, and the final verdict is now in. Led by an intelligent social democrat prime minister (a welder), Sweden had the lowest excess mortality among major European countries during the pandemic, and less than half that of the United States. Sweden’s Covid deaths were below average, and it avoided collateral mortality caused by lockdowns.

Yet on July 29, 2020, the Harvard-edited New England Journal of Medicine published an article by two Harvard professors on whether primary schools should reopen, without even mentioning Sweden. It was like ignoring the placebo control group when evaluating a new pharmaceutical drug. That’s not the path to truth.

That spring, I supported the Swedish approach in op-eds published in my native Sweden, but despite being a Harvard professor, I was unable to publish my thoughts in American media. My attempts to disseminate the Swedish school report on Twitter (now X) put me on the platform’s Trends Blacklist. In August 2020, my op-ed on school closures and Sweden was finally published by CNN—but not the one you’re thinking of. I wrote it in Spanish, and CNN–Español ran it. CNN–English was not interested.

I was not the only public health scientist speaking out against school closures and other unscientific countermeasures. Scott Atlas, an especially brave voice, used scientific articles and facts to challenge the public health advisors in the Trump White House, National Institute of Allergy and Infectious Diseases director Anthony Fauci, National Institutes of Health director Francis Collins, and Covid coordinator Deborah Birx, but to little avail. When 98 of his Stanford faculty colleagues unjustly attacked Atlas in an open letter that did not provide a single example of where he was wrong, I wrote a response in the student-run Stanford Daily to defend him. I ended the letter by pointing out that: