Learning from the neighbors: The diffusion of state broadband policies in the United States

Abstract

This project examines how state broadband policies diffused among the states in the United States over the last 30-year period utilizing a network approach and the State Broadband Explorer dataset curated by the Pew Charitable Trusts’ Broadband Access Initiate. The 621 valid state broadband policies in the U.S. (until January 2021) have been categorized into six main themes: broadband programs, competition and regulation, definitions, funding and financing, infrastructure access, and legislative intent. Our analytical strategy follows a two-step process: (1) to identify the latent network of broadband policy diffusion across the states using the NetInf algorithm; (2) to identify the nodal and dyadic variables that predict the observed diffusion flows. Our objective for the second step is to test out two competing hypotheses: the geographic learning model and the (co-)partisan learning model, which privilege geographic proximity and ideological affiliation respectively as the primary drivers of policy diffusion. The results show that geographic contiguity is the most significant factor predicting broadband policy diffusion. However, the results also identify the low salience of political factors in predicting broadband policy diffusion. Among nodal factors, only one namely divided government (of sender states) is a significant predictor of a diffusion tie. Among dyadic factors, there is one variable that supported political homophily as a significant predictor of diffusion flows (i.e., both states sharing the same type of legislative control). Partisanship appears to be much less of a driver of broadband policy in the U.S. context.