This paper investigates the adoption of national domain name dispute resolution policies (NDRPs) by country code Top Level Domains (ccTLD). We examine the textual similarity between different NDRPs (35 documents) using Exponential Random Graph Modeling (ERGM). First, we measure similarity between the texts (i.e., cosine similarity) and use it to construct a 35*35 adjacency matrix for a directed network early to late adopters. Using EGRM, we predict which nodal attributes and dyadic attributes predict the observed network. We conclude by discussing what the findings imply about the emergence of a global policy regime for the Internet.