Broadband in Schools: Effects on Student Performance and Spillovers for Household Internet Adoption

This work comprises studies on the effects of broadband Internet in schools at three different levels: student performance, household Internet adoption, and individual computer and Internet use patterns and skill acquisition. I focus in the case of Portugal, where by 2006 the Portuguese government had completed a major initiative that upgraded the Internet connection of all public schools, replacing the previously existing connections by broadband. In the first study I focus on the direct effects the introduction of broadband in middle schools had in students' performance. I find that high levels of broadband use in schools are detrimental for 9th grade national exam scores. For the average broadband use in schools, exam scores reduce about 0.97 of a standard deviation from 2005 to 2009. I also find suggestive evidence that the way schools allow students to use the Internet affects students' performance. In particular, students in schools that block access to websites such as YouTube perform relatively better. In the second study, I look at spillover effects of providing broadband to schools in home Internet adoption. I assess the magnitude of these effects using household level data on home Internet penetration and Internet traffic in all schools in Portugal. I find that school broadband use contributes directly to a higher adoption rate in households with children. During 2008 and 2009 school Internet use increased the probability of adopting Internet by 20% in households with children, while no statistically significant effect was found in households without children. In the third study I focus on Information and Communication Technology (ICT) skills and on the dynamics of computer and Internet use inside the household. I provide empirical evidence that the presence of children or young adults in the household does contribute to an increase in the likelihood of having a computer or Internet at home, but does not contribute to an increase in use patterns and skills. Moreover, I find that the presence of children and young adults is associated with lower levels of computer and Internet use and skills.