One Laptop Per Child (OLPC) is a high profile initiative to narrow the inequality of access to ICT and improve educational performance. However, there is little empirical evidence on its impacts. In order to assess the effectiveness of OLPC, we conducted a randomized experiment of OLPC with Chinese characteristics involving 300 third-grade students in Beijing migrant schools. Our results
show that the program improved student computer skills by 0.33 standard deviations and math scores by 0.17 standard deviations.
The program also increased student time spent using educational software and decreased student time spent watching TV. Student selfesteem also improved.