Quantifying Administrative and Functional Border Effects on Commuting and Non-Commuting Flows: A Case Study of the Shanghai-Suzhou-Jiaxing Area

Yige Li, Ying Jiang, Jin DuanISPRS Int. J. Geo Inf 2025 · DOI

TL;DR

This paper investigates how administrative and functional borders impact human movement in the Shanghai-Suzhou-Jiaxing (SSJ) mega-city region. By analyzing mobile phone data, it reveals that both types of borders significantly influence commuting and non-commuting flows, with provincial borders having a particularly strong effect, and non-commuting flows being more sensitive to functional borders.

What Problem Does This Paper Address?

As urban areas expand and regional integration deepens, human movement increasingly transcends traditional administrative boundaries. While studies have explored the "border effect" – how borders influence interactions – there's a gap in understanding the differentiated impacts of both administrative (top-down) and functional (bottom-up) borders on different types of human activity. Existing research often focuses on commuting, neglecting other crucial movements like leisure or shopping, which can reveal different spatial structures. This paper aims to bridge this gap by quantifying and comparing the effects of these two border types on both commuting and non-commuting flows within a major Chinese mega-city region.

Methods

The study employs a multi-step approach:

  1. Delineating Functional Borders:
    • Density-based Aggregation: Grid cells are aggregated into urban clusters based on a synthesized population activity density (combining living, working, and visiting populations). This filters out fragmented areas and focuses on dense urban cores.
  2. Flow-based Delineation: Origin-destination (OD) matrices are created for both commuting and non-commuting flows. The Infomap algorithm, a robust community detection method, is then used on these networks to identify homogeneous groups of spatial units (Functional Urban Communities or FUCs), thereby delineating functional borders.
  3. Functional Characterization: Points of Interest (POI) data is used to understand the functional characteristics (e.g., residential, commercial) of these FUCs.
  4. Centrality Analysis: The PageRank algorithm is applied to rank the importance of FUCs and functional borders.
  5. Quantifying Border Effects:
    • Gravity Model Adaptation: A modified gravity model is used to measure border effects. This model includes dummy variables for administrative borders (provincial, district/county) and functional borders (FUC), as well as variables representing the hierarchical gap between administrative ranks and FUC ranks.

Datasets

  • Mobile Phone Data: Provided by China Unicom, this dataset covers the entire month of June 2023, capturing user locations at a 1 km x 1 km grid resolution. This data is used to infer residence, workplace, and visit places, and to construct commuting and non-commuting flow networks.
  • Points of Interest (POI) Data: Obtained via the Amap API in 2023, this data is used to characterize the functional attributes of the identified FUCs.

Key Insights and Ideas

  • Functional Borders are Key: The study highlights that functional borders, derived from actual human movement patterns, offer a more nuanced understanding of spatial interactions than administrative boundaries alone.
  • Differentiated Flow Sensitivity: Non-commuting flows (e.g., leisure, shopping) are more sensitive to functional borders than commuting flows, suggesting that leisure activities are more constrained by the functional organization of urban space.
  • Provincial Borders as Resources: Contrary to a purely obstructive view, provincial borders in the SSJ area appear to act as catalysts for integration, fostering new cross-border polycentric structures and facilitating economic development.
  • Spatial Heterogeneity: The impact of provincial borders is not uniform. While they generally facilitate flows, their effect is stronger in non-cross-border regions compared to regions already integrated across provincial lines, indicating uneven development.
  • Hierarchical Gaps Matter: The difference in administrative or functional ranks between two locations significantly influences human flows, suggesting that people tend to interact more with places of similar or adjacent ranks.

Key Results

  • Functional Borders Align with Administrative Borders (Partially): Functional borders in peripheral areas of the SSJ region largely align with district/county-level administrative borders, but mismatches occur at city junctions and central urban areas.
  • Cross-Border Polycentric Structure: The SSJ area exhibits a cross-border polycentric hierarchical spatial structure, particularly evident in the non-commuting network, with major and secondary FUCs concentrated in central urban and cross-border regions.
  • Administrative Borders Dominate: Both commuting and non-commuting flows are more influenced by administrative borders than functional borders. Provincial borders have a significant positive effect, effectively reducing geographic distance.
  • Commuting vs. Non-Commuting: District/county borders have a stronger negative effect on commuting flows (people prefer to work near home) but a minimal effect on non-commuting flows (people are willing to travel further for leisure). Functional borders have a greater negative effect on non-commuting flows.
  • Provincial Border Effect: Provincial borders have a significant positive effect, reducing effective distance. This effect is stronger in non-cross-border regions, indicating spatial heterogeneity.

Limitations and Future Work

  • Mobile Phone Data Bias: The study acknowledges limitations due to mobile phone coverage variations across demographic groups, potentially leading to sample bias and skewed estimations of travel flows.
  • Granularity of Non-Commuting Flows: The study could not differentiate specific activities within non-commuting flows (e.g., shopping, healthcare), limiting a fully multi-activity view. Future research could incorporate more granular data for these activities.
  • Explanatory Variables: The border model primarily used population, distance, and border type. Future studies could expand explanatory variables to include functional and morphological characteristics of spatial units for more precise estimations.
  • Multi-Flow Analysis: This study focused solely on human flows. Future research should examine border effects on other flows, such as goods and information, to gain a more comprehensive understanding of regional structures.

Summary updated

BibTeX
@article{li2025quantifying,
  author = {Yige Li and Ying Jiang and Jin Duan},
  title = {{Quantifying Administrative and Functional Border Effects on Commuting and Non-Commuting Flows: A Case Study of the Shanghai-Suzhou-Jiaxing Area}},
  journal = {ISPRS Int. J. Geo Inf.},
  year = {2025},
  doi = {10.3390/ijgi14030133},
}
Figure