Why Small Pockets of Non-Compliance Can Reshape Urban Epidemics

In our work, we investigated how non-compliant behaviour influences epidemic diffusion in urban settings. Public health measures are rarely adopted uniformly: while many individuals follow preventive recommendations, others do not. We wanted to understand how even a relatively small minority of non-compliant individuals might alter the evolution of an epidemic in a city.

To do so, we developed a data-driven modelling framework based on detailed contact networks for three Italian cities—Torino, Milano and Palermo—and combined it with a heterogeneous epidemic model that distinguishes between compliant and non-compliant individuals. Our analysis shows that non-compliance can significantly increase epidemic severity, accelerating the spread and raising the infection peak. We also found that the spatial distribution of non-compliance matters greatly: when non-compliant individuals are geographically clustered, local hotspots can emerge even if city-wide averages appear manageable.

We believe these findings are relevant beyond epidemiology. They show how data science, network modelling and spatial analysis can help reveal system-level risks arising from local behavioural heterogeneity. For policymakers, the message is clear: it is not enough to reduce non-compliance on average; it is equally important to understand where it is concentrated and how urban contact structures amplify its effects.