Simulating a disaster-social solidarity in an emergency

Due to the non-predictive nature of disasters, there is little first-hand data available which in turn is a serious constraint to conduct research on the implication of disasters on urban population. Computational sociology aims at by-passing this issue by simulating the situation that cannot be observed. This paper attempts to use agent-based simulation models to (i) examine the importance of Durkheim's notion of Social Solidarity on behavioral responses in an emergency situation [1], (ii) observe the influence of this Social Solidarity notion in an emergency situation and (iii) validate the use of simulation in this case. Our team created therefore two simulation models representing two hypothetical scenarios, one where the population exhibits low social solidarity, while in the other the virtual entities show high social solidarity. Our findings conclude that low social solidarity leads to behavioral responses of panic and selfishness while high solidarity leads to an ordered and organized evacuation. With these findings, a theory of how a typical urban population would react to a large-scale disaster is formulated. The validity and effectiveness of agent-based modeling as an epistemological tool is also discussed.