Two New Studies Say Masks and Distancing Didn't Save Lives. They Make Critical Errors.

A public-health author writing in STAT News is challenging two recent reanalyses of pandemic-era nonpharmaceutical interventions, or NPIs, the catch-all term for measures like masking, distancing, and school closures that were used before Covid vaccines existed. The argument is targeted, not generic: both studies, a Princeton political scientists' book by Stephen Macedo and Frances Lee and [a Lancet paper by Tom Bollyky and his Institute for Health Metrics and Evaluation (IHME) colleagues], which the author describes as containing critical methodological errors that bias their findings against detecting any benefit of those measures.

The author anchors the counter-case in a single ratio: roughly one life saved per 150 infections prevented during the pre-vaccine period. The critique matters because the two studies are positioning themselves as the corrective record for the public's pandemic memory, and the next respiratory virus will be met with whatever default the current "settled" record has produced.

The Princeton book, "In Covid's Wake: How Our Politics Failed Us," is a work of political history that questions the policy response. The Lancet paper is the global-health version. Led by Bollyky at IHME, it models what would have happened without those measures and concludes they had limited effect on cumulative death counts. The STAT opinion, dated July 21, 2026 does not dispute the urgency of asking whether the response overshot. It disputes the design choices that produced a "no benefit" finding.

The author frames the methodological problem as one that disproportionately hits reanalyses rather than the original pandemic-era studies. The original work generally used within-jurisdiction time-series methods, contact-survey data, or synthetic-control designs that absorbed some of the noise from voluntary behavior change and patchy compliance. Reanalyses that pool data across many jurisdictions and treat the presence of a mandate as the only variable of interest lose that absorption. The result, the author argues, is a finding of near-zero effect that reads as a measurement artifact rather than a true null.

The stakes the author names are not 2020 or 2021. They are the next pandemic. Respiratory viruses with pandemic potential are now a standing category on the global health risk register, and the playbook for the first sixty days of any new outbreak, the period before vaccines, will be set by what policy makers, school districts, and hospital systems believe worked last time. A "settled" finding that masks, distancing, and school closures did not save lives does not just rewrite history. It strips the default response of its main tools before the next pathogen arrives.

The Princeton book is a work of political history and was not designed to settle a quantitative question; the Lancet paper is peer-reviewed and was. The author of the STAT opinion is asking the global-health community to treat both as preliminary until the methodological questions are answered, rather than as the closing entry in the Covid record. Readers who want the policy-relevant takeaway can hold a single claim in mind: the per-infection mortality benefit of pre-vaccine NPIs was small per person and large in aggregate, and any reanalysis that finds zero is, on its face, suspect until the design choices are examined.

Watch item: the IHME group's response to the specific methodological critiques, and whether the Macedo and Lee book generates a quantitative reply from epidemiologists rather than another political-history treatment. Both will shape the version of the Covid record that gets taught to the people who will run the next response.