Public libraries and residential broadband adoption: Do more computers lead to higher rates?

•Do libraries with publicly available Internet connections influence local rates of residential broadband adoption?•The presence of a library is positively associated with local broadband adoption rates, but only in the most rural counties.•A propensity score matching technique shows that this relationship is not necessarily causal, however.•An increased number of library computers (or hours of use) did not lead to higher residential broadband adoption rates.•Evaluating the impact of public libraries using data external to the libraries themselves is vital.

Providing access to computers with high-speed Internet connectivity is a central mission of public libraries in the United States. One pertinent, currently unanswered question is whether library Internet access leads to increasing residential broadband adoption rates in the communities that the libraries serve. This paper uses simple county-level regression analysis to document a positive association between a higher number of libraries and household broadband adoption rates as of 2013—but only in rural areas. This correlation does not imply causation, however. A propensity score matching technique is used to demonstrate that counties with libraries that aggressively increased their number of Internet-accessible computers between 2008 and 2012 did not see measurably higher increases in their rates of residential broadband adoption. These findings lend themselves to future research questions including how to appropriately measure broadband ‘adoption’ outside of the home and methods for engaging library patrons that ultimately encourage residential broadband adoption.