Cyber versus Brick and Mortar: Achievement, Attainment, and Postsecondary Outcomes in Pennsylvania Charter High Schools

Pennsylvania statewide administrative records on charter high-school entrants

Pennsylvania students entering brick-and-mortar or cyber charter high schools

Statewide administrative data covering attendance, achievement, graduation and postsecondary outcomes, comparing brick-and-mortar with cyber charters. Cyber enrolment is chosen, not assigned, and the author frames the cyber estimates as associations; the value here is that it is the largest study of a fully home-delivered secondary education with hard attainment outcomes.

Brick-and-mortar charters show no or positive effects across outcomes, concentrated among urban, Black and economically disadvantaged students. Cyber charters are "almost universally worse", with little heterogeneity: students enrolling in a cyber charter at the start of ninth grade are 9.5 percentage points less likely to graduate on time, 16.8 pp less likely to enrol in college, and 15.2 pp less likely to persist beyond one semester of postsecondary study.

Medium. Selection into cyber charters is not random and plausibly captures already-struggling students, which would inflate the negative estimate; the contrast with brick-and-mortar charters drawn from an overlapping choice population partially bounds this, and the effect sizes are far larger than plausible selection on observables.

The direction replicates Fitzpatrick et al. 2020's within-student Indiana finding and the CREDO online-charter results; three independent state datasets, three methodologies, same sign and same rough magnitude class.