New research · 2024
Urban vibrancy · 7 Italian cities · Mobile phone data
The data
Modern cities generate continuous streams of digital trace data. Call Detail Records — logged every time someone makes or receives a phone call — let researchers map human presence across an entire city at 15-minute resolution. This study uses that data to ask: do men and women move through and inhabit urban spaces differently?
Call Detail Records from Telecom Italia — Italy's largest mobile operator. ~2 months of activity (March–April 2015), captured at 15-minute intervals across 7 cities.
Each city is divided into a spatial grid. Male and female phone activity is counted per cell, per time window. Results validated against census data: Kendall τ = 0.55–0.72.
Gender difference Δ = Male activity − Female activity. Positive means more men present; negative means more women; zero means equal use.
7 cities studied
Key findings
Across all seven cities, male and female activity patterns track broadly similar rhythms — but they diverge. The gap is not uniform across the day, and — critically — it is not uniform across space. Some parts of a city are consistently more male-dominated; others, more female.
"Significant gender differences in urban vibrancy exist across all 7 cities studied, with evidence of spatial clustering."
The gap isn't just about when — it's about where
Core result
Two urban features drive gender differences in opposite directions — and distinguishing between them is crucial for policy. It is not commercial density in general that creates segregation, but specifically the presence of informal social spaces.
cafés · pubs · parks · community spaces
↑ Larger gender gapAreas with higher density of third places — informal social spaces that are neither home nor work — show significantly larger male-female differences in urban activity. These spaces, in theory open to all, appear to be used unequally across genders.
5 CATEGORIES EXAMINED
shops · services · amenities · transport
↓ Smaller gender gapAreas with higher overall density of Points of Interest (all types combined) show smaller male-female differences in urban activity. General commercial and civic density is used more equally across genders.
CONSISTENT ACROSS
Spatial dynamics
A café district does not only affect the block it sits on. Gender-use patterns in one area bleed into neighbouring grid cells — a spatial spillover that conventional regression models would miss entirely. The study uses Spatial Autoregressive models (SAR) to capture this.
Click or hover the grid to see spillover effects
Changes in one area's gender gap ripple into neighbouring grid cells — the indirect effects captured here were consistent and significant across the three largest cities and the pooled model.
Additional findings
The gender gap in urban use is not an artefact of a particular time of day, type of week, or city size. It is a structural feature of urban space.
Time-stable — No strong differences between weekday and weekend, or between morning, afternoon, evening and night. The gender gap in how cities are used is a persistent structural feature, not a time-of-day effect.
Density matters more than diversity — The diversity of urban features was generally not significant; density was. This contrasts with earlier work on age groups, where diversity played a larger role.
Smaller cities, sharper effects — In Milan and Rome the associations are present but more diffuse. In smaller cities such as Bari and Palermo the relationship between urban features and gender differences is stronger — possibly because residents have fewer alternative spaces to choose from.
Implications
Understanding where and how gender segregation manifests in cities — beyond where people live and work — is essential for designing fairer, more equitable urban environments. Streets, parks, and cafés that appear neutral on a planning map may in practice be used very differently by men and women.
This research demonstrates that mobile phone data, combined with open geographic data from OpenStreetMap, can reveal gender inequalities in urban space at a resolution and temporal frequency that is simply impossible to achieve with traditional survey or census methods. As mobile data becomes more widely available and privacy-preserving methods mature, this approach opens a new window into the lived experience of cities — across genders, time of day, and place.
Collins, T., Di Clemente, R., Gutiérrez-Roig, M., & Botta, F. (2024). Spatiotemporal gender differences in urban vibrancy. Environment and Planning B: Urban Analytics and City Science, 51(7), 1430–1446.
doi.org/10.1177/23998083231209073