This paper provides empirical evidence on the effects of broadband Internet usage in schools on student performance in terms of national exams scores. We use a rich panel of data that has information on test scores, as well as broadband usage for all schools in Portugal, allowing us to control for school-specific effects. Additionally we use an instrument to account for possible unobserved time-varying effects. For 9th grade students, our estimates indicate that a higher use of broadband Internet is detrimental for students' test scores, despite this effect seeming to be wearing off with time. We find that the adverse effect tends to be reinforced for boys and weakened for girls, compared to the pooled estimates. We also find that schools with worse performance right before the introduction of broadband Internet in schools suffered the most.

Our results suggest that introduction of broadband Internet in schools is not enough to improve students' performance. Broadband deployment in schools needs to be accompanied by complementary measures that support the use of the technology in productive ways.