With Universal Service for All: Broadband Internet Access and Adoption in Rural High-Cost Areas

Abstract

This thesis poses two research questions surrounding the rural-urban digital divide in broadband services, namely (1) how the presence and magnitude of funding to support broadband provision in rural high-cost areas affects broadband access in these areas, and (2) how local levels of social capital in rural communities may affect the degree to which broadband services are adopted with that community. This research distinguishes itself from the existing literature by focusing on more recent disbursements of funding, and by proposing and testing a model for broadband adoption that accounts for the potential role of social capital in influencing broadband usage. It uses county-level data on broadband access, adoption, funding, social capital, ruralness, and other variables to test a number of hypotheses about the funding-access relationship and the social capital-adoption relationship. Multivariate linear regression analysis is used to produce coefficient estimates for each of the explanatory variables, and these estimates are used to observe the nature and magnitude of funding and social capital on their respective response variables. The results of the thesis indicate that, regarding funding, the presence of funding is slightly positively associated with greater broadband access improvements in the middle range of counties, though the mean change for funded counties is lower, and that funding amounts are somewhat negatively correlated with improvements in broadband access. Funding is also found to have an inconsistent interaction with the ruralness of a county depending on the sample of areas considered. Regarding adoption, it is observed through proxy variables that bridging social capital is positively correlated with broadband adoption in both rural and urban areas, and bonding social capital is somewhat, though not consistently significantly, negatively correlated with broadband adoption. Social capital in general is also found to interact generally negatively with other socioeconomic and demographic correlators in its effect on broadband adoption. Possible explanations for the observed relationships are considered and discussed, along with the policy implications of these findings and potential avenues for further research.