This study tests 2 competing theoretical models that attempt to understand the roles that broadband Internet plays in society. The linear model posits that the gains or harms introduced by the Internet via narrowband will further increase with the adoption of broadband. On the contrary, the differential gains model proposes that changes stemming from advances to different stages of Internet connection technology should be unique to each technological advance. Findings show that support for these models was contingent upon the domain that each criterion variable represents. For hard engagement, such as political discussion, hard knowledge, and civic participation, there was support for the differential gains model, but for soft engagement, such as informal socializing and soft knowledge, the linear model was supported.