Effects of mobility restrictions during COVID19 in Italy

TL;DR

This study examines how COVID-19 mobility restrictions in Italy affected the mobility network, revealing deeper network damage than traditional metrics and proposing a method to estimate economic indicators from mobility data.

Contribution

The paper introduces a network science approach to assess mobility network damage during lockdown and proposes a novel method to estimate economic metrics from mobility data.

Findings

Mobility network was more damaged during lockdown than node- and edge-level metrics suggest.

Main Italian Provinces experienced similar effects despite differences.

Proposed a method to estimate real-time Province GDP from mobility data.

Abstract

To reduce the spread and the effect of the COVID-19 global pandemic, non-pharmaceutical interventions (NPIs) have been adopted on multiple occasions by governments. In particular lockdown policies, i.e., generalized mobility restrictions, have been employed to fight the first wave of the pandemic. We analyze data reflecting mobility levels over time in Italy before, during and after the national lockdown, in order to assess some direct and indirect effects. By applying methodologies based on percolation and network science approaches, we find that the typical network characteristics, while very revealing, do not tell the whole story. In particular, the Italian mobility network during lockdown has been damaged much more than node- and edge-level metrics indicate. Additionally, many of the main Provinces of Italy are affected by the lockdown in a surprisingly similar fashion, despite their geographical and economic dissimilarity. Based on our findings we offer an approach to estimate unavailable high-resolution economic dimensions, such as real time Province-level GDP, based on easily measurable mobility information.