We use social network analysis to better understand historic data on the administration of local governments. Despite advances in e-government applications, the public sector lags behind in analytics because information is locked in legacy data formats. Can e-government researchers bridge the gap between legacy data and analytics? We argue that computational analytic methods, popular in big data applications, can explain patterns that have gone unquestioned in previous research on government. Specifically, we consider how explanations of state government authority can be explained using a network perspective. These data were originally designed to describe administrative differences US territories and states. We investigate methodological challenges in building a weighted social network to confirm existing measures for calculating the power of the state governor. This project reports on the initial step in a broader study to cover all 50 states across multiple years and agencies. We compare the states that experienced the greatest differences in gubernatorial appointment power between 1992 and 2012 Texas and Massachusetts. In addition, we identified that Information Systems agencies moved closer to gubernatorial control across all 50 states. Social network analysis improves existing measurements because it indicates the relationship between the governor and other top officials and agencies. An analytics approach explained where the power shifted across states and across time. Computational analysis of existing government data matches findings from previous studies as well as adding additional explanatory power.