Tracking riots with models of disease spread
Tracking riots with models of disease spread lead image
Epidemiological models have often been used to study rioting behavior, as the aggregated disorder follows the same pattern as the spread of a contagious disease. At first, the number of infected (or disruptive) individuals grows quickly. It reaches a maximum, then decays exponentially as the population heals or disperses. Such models have shown excellent agreement with multiple real-world riots.
Carlos Cartes applied an epidemiological model to try to mimic the Chilean riots of 2019, which were sparked by an increase in transportation fare and a desire for better living standards.
Riots like those in Santiago can cascade into violence and damage public and private property. In Chile, the 2019 unrest cost about 1.1% of the country’s Gross Domestic Product.
“Therefore, the ability to predict the temporal evolution of a potentially violent crowd and the possibility of controlling it — or at least mitigating its actions — is a subject of substantial practical importance for any authority involved in politics or security,” said Cartes.
Population, income distribution, geographical distances, and an initial triggering event are all included in the model. The algorithm then follows the disorder intensity as it evolves in time and space. Cartes included Santiago’s subway network influence on the model by considering two areas connected by the subway “closer” than otherwise.
“The model adequately reflected high activity in Santiago’s most deprived areas and increased intensity due to the subway network’s influence,” he said. “Nevertheless, it could not replicate the very intense riots in Santiago’s better-connected regions.”
To describe the riots more accurately, the model needs a mechanism to displace the population between different areas of the city. Cartes plans to explore this in future work.
Source: “Mathematical modeling of the Chilean riots of 2019, an epidemiological non-local approach,” by Carlos Cartes, Chaos (2022). The article can be accessed at https://doi.org/10.1063/5.0116750 .
This paper is part of the Complex Systems and Inter/Transdisciplinary Research Collection, learn more here .
