Concerns over privacy on social networking sites generally focus on the options available to individual users for protecting personal information on their own pages. However, similar information is often exposed on users’ friends’ pages in the form of public posts, comments or photographs. Users’ actual privacy depends not just on their own settings but on their friends’ settings as well.
In this paper I define a computational model for the exposure of personal information on a social networking site and use it to analyze the importance of aggregate privacy tendencies for individual privacy. I find that privacy settings have a spillover effect, and that exposure is further affected by network topology.