To What Extent Does In-Person Schooling Contribute to the Spread of COVID-19? Evidence from Michigan and Washington

In this paper we use data from two states—Michigan and Washington—on COVID case rates at the county level linked to information on the type of instructional modality offered by local public school districts to assess the relationship between modality and COVID outcomes. We focus primarily on COVID case rates, but also provide estimates for hospitalizations (in Washington only) and deaths. Our preferred district and month fixed effects models exploit within district (over time) variation in instructional modality and account for time-invariant district factors. In both states, we find evidence that instructional modality does lead to increases in COVID spread in communities with moderate to high levels of pre-existing COVID cases, although the causal effect is small in magnitude.

Working Paper No. 247-1220 was originally released in December 2020 and has since been updated to Working Paper No. 247-1220-3, released in July 2021.