Dynamic Network Analysis via Diffusion Multipliers

Research output: Working paperPreprint


We develop new methods to understand the impact of shocks or policy interventions across a dynamic network, allowing the data to speak about the connectedness of nodes through the system diffusion multipliers. These are derived from a set of unusually heterogeneous spatial/temporal parameters, for which we show that estimates consistent at the required parametric rate can be easily found. Our techniques are applied to a novel dataset of civilian casualties due to armed violence in post-invasion Iraq, where we find evidence consistent with the use of violence as a means to maintain influence in the nascent political institutions as well in reprisal for past acts.
Original languageEnglish
Publication statusUnpublished - Sept 2021

Cite this