Drawing on Bachrach and Baratz's first and second faces of interest group power, we explore the relationship between teachers' union power and reopening decisions during the fall 2020 semester in 250 large districts around the United States. We leverage a self-collected panel data set of reopening decisions coupled with measures of teachers' union first face power (drawn from social media postings on teachers' unions' Facebook pages) and second face power (operationalized as district size, whether the school district negotiates a collective bargaining agreement with the teachers' union, the length of the collective bargaining agreement, and the amount of revenue raised by the union). We found that school districts where teachers' unions exhibit strong second face power (but not first face power) were less likely to start the school year with in-person instruction, were less likely to ever open during fall semester with in-person instruction and spent fewer weeks in in-person learning.