Intra-Household Peer Effects in Smartphone Adoption

This paper examines intra-household peer effects in the adoption of smartphones using unique South Korean panel data. Consistent estimation of peer effects in this setting needs to address two key challenges: homophily and endogenous sample attrition. We address both challenges and obtain consistent estimates of peer effects, by first-differencing the individual-level panel data and then using longer differences of the independent variables as instruments. The estimation results show that an individual becomes much more likely to adopt a smartphone if other household members have previously adopted one. The analysis also reveals that failure to account for endogenous attrition of individuals after product adoption would lead to a significant under-estimation of peer effects.