Background : Education, like other service operations, depends on the quality of its delivery channel. Broadband has become a critical delivery channel for learning in the current education system, but access and speed remain uneven across U.S. communities. Aim : Scholars disagree on how broadband access affects educational outcomes, and prior work often overlooks broadband speed. We extend prior work by assessing how both broadband access and speed together influence U.S. math proficiency. Methods : We analyze a national panel of U.S. school districts from 2017 to 2022 and link broadband measures to student proficiency with a dynamic panel model. Results : School districts with greater average access and smaller within-district gaps tend to have higher math proficiency. Nationally, a one percentage-point increase in broadband access corresponds to a 0.545 percentage-point increase in math proficiency, a 1.22% increase relative to the average baseline. We find speed also matters: gains rise with faster available connections, then reverse at very high speeds. Conclusion : Access alone is not enough. Our study shows that broadband policy should move beyond “more is better” and consider broadband speeds that maximize learning. As one headmaster noted: “The internet is essential, but proceed with caution!”
Niewoehner et al. (2026) studied this question.