Zimmerman et al. conducted a time-series analysis to study the impact of school closures in the spring of 2020 along with other policy actions (non-essential business closures, bans on large public gatherings, stay-at-home orders, and restaurant closures) and proxies for behavioral change (work reduction, increase in time at home, searches for hand sanitizer, and reductions in restaurant meals). The authors found that significant behavior change occurred prior to school closures. When only controlling for state-level covariates, each day of earlier school closure was associated with a 14.6% reduction in COVID-19 incidence. However, when also including days since gathering ban (policy action) and reduction in time spent at work (behavioral change), each day of earlier school closure was associated with a 3.5% reduction in incidence. Meanwhile, each day of earlier behavioral change was associated with a 9.3% reduction. The simulations also suggest that a two-week delay in school closures would have resulted in an additional 587,000 cases, whereas a two-week delay in behavioral change would have resulted in an additional 4.3 million cases. Collectively, these findings indicate that while school closures did prevent COVID-19 cases and deaths, they were not as effective of a public health measure as voluntary behavioral changes.