Balancing Efficiency and Vulnerability in Social
Networks: A Simulation Study
Network centrality is thought to be crucial for the efficient
distribution of information or resources through a social network. Inherent
to this structure, the flow of information in high-centrality networks can
be easily undermined. By introducting two measures, network efficiency
and vulnerability, we show that efficiency is compromised more in high centrality
networks than in other, e.g., low centrality, networks following the removal
of network positions. Vulnerability was measured as the loss in efficiency
resulting from the elimination of nodes. The resiliency of a network
was then assessed following progressive removal of nodes and all adjacent
nodes. Nodes were removed in two ways: 1) randomly, which modeled incidental
vulnerability; or, 2) based on the calculated centrality measures, which
modeled vulnerability to a strategic attack. Our findings highlight
the social impact of removing a person from a network, which is likely to
disable, or make inactive, all direct ties to that individual. Such
network analyses can be used to undermine or fortify existing network structures
necessary for law enforcement and those concerned with national security.
Phillip Bonacich
University of California in Los Angeles
Department of Sociology
bonacich@soc.ucla.edu
Elisa Jayne Bienenstock
University of California in Irvine
Department of Sociology
ejb@uci.edu