Source-linked AI summary
Evidence of economic segregation from mobility lockdown during COVID-19 epidemic
Giovanni Bonaccorsi, Francesco Pierri, Matteo Cinelli, Francesco Porcelli, Alessandro Galeazzi, Andrea Flori, Ana Lucia Schmidt, Carlo Michele Valensise, Antonio Scala, Walter Quattrociocchi, Fabio Pammolli
TL;DR
The paper asks how COVID-19 lockdown restrictions affect economic conditions across individuals and local governments. Using near-real-time Italian mobility data and modeling mobility change as an exogenous shock, it finds stronger reductions in municipalities with lower individual income, higher inequality, and higher fiscal capacity.
Problem
The paper examines how lockdown restrictions affect economic conditions and whether their effects are uneven across individuals and municipalities.
Method
The study analyzes near-real-time Facebook mobility data for Italian municipalities and models mobility change as an exogenous shock similar to a natural disaster.
Results
Mobility and connectivity reductions are stronger for municipalities with low average individual income, high income inequality, and higher fiscal capacity.
Takeaways & Limitations
The lockdown has an asymmetric impact, affecting poorer individuals within municipalities with strong fiscal capacity.
Abstract
from arXiv · showhide
In response to the COVID-19 pandemic, National governments have applied lockdown restrictions to reduce the infection rate. We perform a massive analysis on near real-time Italian data provided by Facebook to investigate how lockdown strategies affect economic conditions of individuals and local governments. We model the change in mobility as an exogenous shock similar to a natural disaster. We identify two ways through which mobility restrictions affect Italian citizens. First, we find that the impact of lockdown is stronger in municipalities with higher fiscal capacity. Second, we find a segregation effect, since mobility restrictions are stronger in municipalities for which inequality is higher and where individuals have lower income per capita.
5 Department of Information Engineering, Universit di Brescia,
The passage identifies an affiliation with the Department of Information Engineering at the University of Brescia.
- The affiliation is the Department of Information Engineering, Università di Brescia.
- The department is listed as an institutional affiliation rather than a research finding.
- No methods, results, or conclusions are stated in this passage.
7 Department of Physics, Politecnico di Milano
The passages identify an affiliation with the Politecnico di Milano and provide a corresponding-author contact line.
- The affiliation is CADS, Joint Center for Analysis, Decisions and Society, Human Technopole, Politecnico di Milano.
- No research methods or findings are provided in these passages.
- The passage also includes a corresponding-author contact instruction.
Introduction
The paper examines the economic effects of COVID-19 mobility restrictions in Italian municipalities, treating mobility change as an exogenous shock. It reports stronger reductions in higher-fiscal-capacity municipalities and greater contractions in lower-income and more unequal municipalities.
- The study investigates how lockdown restrictions affect economic conditions across Italian municipalities.It focuses on identifying the economic conditions of the most and least affected areas.
- The authors model mobility change as an exogenous shock analogous to a large-scale natural disaster.The comparison motivates an empirical framework for analyzing lockdown-related economic disruption.
- Facebook near-real-time mobility data and official economic statistics are used at municipal resolution.Mobility variation serves as a proxy for economic downturn and restrictions affecting population movement.
- More than 90%: mobility trends fell in Italy after lockdown across retail, tourism, and services.The workplace mobility disruption supports using mobility flows as a proxy for economic damages.
- Lockdown-induced mobility reduction is stronger in municipalities with higher fiscal capacity, lower per-capita income, and higher inequality.The introduction frames these patterns as evidence of uneven economic effects and highlights the need to compensate lost local fiscal capacity.
Results and Discussion
The paper analyzes lockdown-driven changes in Italy’s municipal mobility network and their relationship with economic characteristics. It finds network fragmentation and an asymmetric burden: poorer individuals and more unequal municipalities experience stronger mobility contraction, while municipalities with higher fiscal capacity are also more affected.
- Mobility network: After 21 days of national lockdown, the mobility network shows striking fragmentation and its peripheral municipalities are most affected.
- Mobility network: The number of weakly connected components increases while the largest connected component decreases, confirming the breakdown of mobility hubs.
- Mobility network: Network efficiency decreases drastically, indicating a pronounced drop in the mobility potential of the network.Efficiency combines network cohesiveness and node distances to measure how efficiently individuals or information travel.
- Economic relationships: Quantile regression examines changes in Nodal Efficiency across the distribution rather than conditioning only on the mean.This approach targets dynamics at the tails, where effects may otherwise appear insignificant under linear methods.
- Economic relationships: Lower-income municipalities experience stronger mobility and connectivity reductions, while high-income municipalities experience less intense changes at the lower end of the distribution.The asymmetric joint distribution supports a possible segregation effect in which low-income individuals bear greater lockdown-related economic consequences.
- Economic relationships: Municipalities with higher fiscal capacity and lower aggregate deprivation experience stronger losses in mobility efficiency.
- Economic relationships: Higher inequality is significantly associated with stronger mobility contraction at the lower end of the mobility-reduction distribution.The findings link stronger mobility changes with both low income and high inequality.
- Conclusions: The evidence indicates an asymmetric impact in which poor individuals within municipalities with strong fiscal capacity are most affected.The authors also report stronger effects in municipalities with more buildings per capita, corresponding to lower urban density.
Conclusions
The analysis links lockdown-related mobility and connectivity reductions to municipalities’ economic characteristics. The effects are uneven, affecting poorer populations and municipalities with higher income inequality, while reductions are also stronger in municipalities with greater fiscal capacity.
- Lockdown-related reductions in connectivity tend to be stronger in municipalities with lower average individual income and higher income inequality.
- The reduction in connectivity tends to be higher for municipalities with higher Fiscal Capacity.
- The lockdown seems to unevenly affect the poorer fraction of the population.
- Municipalities where income inequality is greater experienced stronger changes in mobility.
- The findings suggest asymmetric fiscal measures, including grants for poor people and compensation for rich municipalities’ loss of fiscal capacity.
Figures
The figures show that Italy’s mobility network became less connected and efficient after lockdown, while larger mobility reductions were associated with local economic conditions. Quantile analysis examines how income, inequality, fiscal capacity, and deprivation relate to these changes across the distribution.
- The mobility network’s connectivity changed after lockdown, with weakly connected components increasing and the giant connected component decreasing over time.Figure 1 tracks daily network snapshots and reports significant trends for both measures.
- Global efficiency of the Italian mobility network significantly decreased from February 23rd to April 4th.Efficiency uses reciprocal edge weights to represent distances between municipalities.
- −0.153 Pearson correlation links relative nodal-efficiency change with the Deprivation Index, while 0.263 links it with Irpef per Capita.The figure reports negative correlations for deprivation and positive correlations for income per capita, with some variation across time periods and specifications.
- Table 1: Quantile regression estimates relative efficiency changes against income per capita while controlling for deprivation, fiscal capacity, real-estate resources, inequality, and regional factors.The extended model uses standardized variables, bootstrap standard errors, and 2,345 observations.
- Figure 4: Figure 4 compares quantile-regression coefficients and 95% bootstrap confidence intervals with OLS estimates for income, inequality, fiscal capacity, and deprivation.The coefficients are plotted across percentiles of relative nodal-efficiency change.