Document Type


Publication Date



We conceptualize students’ opportunities to learn remotely during the initial school closures associated with the COVID-19 pandemic. We then examine variation in remote instruction using an original statewide survey of teachers in Tennessee, deployed just a few weeks into the closures. Using three-level logistic regression models, we explore potential predictors of regular remote instruction, including prepandemic measures of broadband access, the demographic composition of schools, and measures of district policy responses created from districts’ public COVID-19 plans. We find that teachers in schools serving more economically disadvantaged students and in rural districts are less likely to report regular remote instruction, especially via providing digital resources and holding virtual classes or tutoring. Fewer opportunities for Tennessee’s rural students appear driven in part by lower community access to high-speed broadband, but district policies to distribute technology may partially mitigate this gap in access.