Measuring External Stress in Political Committee Networks

With the increasing abundance of temporal data, recent research in social network analysis has explored the impact of external events or “shocks” on the structure and evolution of the underlying graph. This type of analysis stands in contrast to analysis of network dynamics based on properties intrinsic to the graph itself. An understanding of how social networks are shaped by external events can provide insight into understanding networks as dynamic objects capturing complex relationships and as a new tool for modeling human behavior. In this project, we seek to measure the impact of external stress in a network of donors to candidates and political action committees (PACs) during the 2016 election cycle. In addition, we propose several null models and associated stress disruption algorithms to develop a theoretical framework for network responses to external effects based on our observed results The 2016 U.S. Presidential election cycle, notable for several news events that shifted polls and public opinion considerably, struck us as a powerful case study into understanding the interplay between the complex monetary relationships defining political campaigns. Given the recency and historical significance of this election, we ultimately hope that our work can provide an informative analysis of the behavior of donors during the election while, on a more abstract level, providing insight into responses to stress in empirical networks.