The Impact of Social Segregation on Human Mobility in Developing and Urbanized Regions

This study leverages mobile phone data to analyze human mobility patterns in developing countries, especially in comparison to more industrialized countries. Developing regions, such as the Ivory Coast, are marked by a number of factors that may influence mobility, such as less infrastructural coverage and maturity, less economic resources and stability, and in some cases, more cultural and language-based diversity. By comparing mobile phone data collected from the Ivory Coast to similar data collected in Portugal, we are able to highlight both qualitative and quantitative differences in mobility patterns - such as differences in likelihood to travel, as well as in the time required to travel - that are relevant to consideration on policy, infrastructure, and economic development. Our study illustrates how cultural and linguistic diversity in developing regions (such as Ivory Coast) can present challenges to mobility models that perform well and were conceptualized in less culturally diverse regions. Finally, we address these challenges by proposing novel techniques to assess the strength of borders in a regional partitioning scheme and to quantify the impact of border strength on mobility model accuracy.

💡 Research Summary

The paper investigates how social segregation, cultural and linguistic diversity, and infrastructural differences shape human mobility patterns in a developing country (the Ivory Coast) compared with a highly industrialized nation (Portugal). Using anonymized Call Detail Records (CDRs) from the Data for Development (D4D) challenge for the Ivory Coast (SET1 and SET2) and a comparable Orange dataset for Portugal (SET3), the authors analyze over 2.5 billion calls and 400 million records spanning 150 days (Ivory Coast) and two years (Portugal). They also incorporate high‑resolution population density maps derived from census data, assigning each mobile antenna a Voronoi cell and the corresponding population.

First, the authors compute the probability density function of individual travel distances (Δr) and fit a truncated power‑law model P(Δr) = (Δr + Δr₀)⁻ᵝ exp(‑Δr/κ). Both countries exhibit similar cutoff distances (κ ≈ 110 km), but the Ivory Coast shows a larger exponent (β ≈ 1.62 vs. 1.37 for Portugal), indicating a faster decay of travel probability with distance. When the data are partitioned by first‑level administrative units (regions in Ivory Coast, districts in Portugal), the Ivory Coast displays markedly higher variability in β and κ across regions, whereas Portugal’s parameters are relatively homogeneous. This suggests that intra‑country social and linguistic borders strongly modulate mobility in the developing context.

Second, the radius of gyration (r_g) distributions are examined. Both nations follow a scale‑free pattern, confirming that overall mobility magnitude is comparable despite differing socioeconomic conditions.

Third, daily commuting behavior is explored by measuring the fraction of inter‑call events that involve a displacement within a 40‑minute window, focusing on weekdays. Portugal’s population shows a higher overall probability of displacement, with a pronounced morning peak (7–9 am) and a smoother evening decline. In the Ivory Coast, the capital Abidjan exhibits a higher mobility probability than the rest of the country, but the national average is lower than Portugal’s. Mean inter‑event travel distances are also larger in Portugal, especially in Lisbon, where a sharp increase occurs during the morning commute; this spike is absent in Abidjan, implying that many Ivorian commuters live and work within close proximity.

Fourth, the authors stratify commuting by travel‑distance bins (0‑1 km, 1‑5 km, 5‑10 km, 10‑20 km, 20‑50 km). In the Ivory Coast, all bins display a bimodal pattern (morning and evening peaks) with a deep valley between them that deepens for longer trips, reflecting limited mid‑day movement for long‑distance commuters. In Portugal, short‑distance bins reveal an additional midday peak around 13:00, likely linked to mandatory lunch breaks enforced by labor regulations, a feature absent in the Ivory Coast data.

Fifth, the paper constructs mobility networks where nodes are antennas and edges represent aggregated movements. Using modularity (Q) as a measure of community strength, the Ivory Coast network exhibits higher modularity, indicating dense intra‑regional connections and sparse inter‑regional flows. This aligns with the hypothesis that cultural and linguistic boundaries act as “soft borders” that restrict movement. Portugal’s lower modularity reflects a more integrated national mobility fabric.

Finally, the authors propose a novel metric called “border strength” to quantify how strongly administrative or cultural borders impede mobility. By comparing the density of edges crossing community boundaries with the expected density under a random null model, they demonstrate that higher border strength correlates with larger prediction errors in standard mobility models (e.g., gravity or radiation models). Consequently, models calibrated on data from industrialized nations may systematically overestimate mobility in regions with pronounced social segregation.

Overall, the study provides a methodological framework for incorporating social segregation into mobility modeling, highlights the limitations of transferring models across development contexts, and offers actionable insights for policymakers, urban planners, and service providers seeking to design infrastructure and interventions that respect the nuanced mobility realities of developing societies.