Year of Graduation


Level of Access

Restricted Access Thesis

Embargo Period


Department or Program


First Advisor

Daniel Stone


Instant streaming has transformed what it means to be a viewer. Armed with the choice of not only what, but how to watch a television show, the modern viewer can indicate their preferences for television through their viewing history. In this study, I create and use an innovative data set of Bowdoin College student Netflix viewing histories. I construct measures of popularity and quality of television shows within the sample at the season level, and estimate OLS regressions using the measures and other observable factors of a series. I find that whether a show is a Netflix original, whether the cast is famous, and how highly previous audience viewers have rated the show all positively affect a season’s popularity in this sample. I also find evidence that quality as measured by the episode completion rate of season does contribute to a show’s season level popularity.


Available only to users on the Bowdoin campus.