How Far to Go to Encounter the Differences: Examining the Nonlinear Relationships Between Low-Income Migrants’ Activity-Space Segregation and Traveling Distances in Shenzhen, China

Sociospatial segregation has long been a critical subject in urban studies. In recent decades, increasing attention has been given to the segregation experienced in activity spaces beyond the well-examined residential locations. A large body of studies revealed that disadvantaged groups (e.g., low-income migrants) can access more opportunities for cross-group interaction by engaging in activities outside of their residential areas. This study extends this body of research by examining whether individuals have to travel a certain distance before their activity-space exposure noticeably increases (i.e., gaining more opportunities for cross-group interaction), with a focus on low-income migrants in Shenzhen, China. Through analyzing the nonlinear relationship between travel distance and activity-space exposure using cellphone data and gradient boosting decision tree (GBDT) models, the study not only confirms the possibility of increased interactions in activity spaces, but also reveals that low-income migrants must travel a threshold distance before their activity-space exposure significantly increases. For central-city migrants, this threshold is 2 km for work activities or 3 km for nonwork activities, whereas suburban migrants must travel at least 9 km for work activities or 8 km for nonwork activities. These results highlight disparities between urban locations: Central-city areas offer shorter distance opportunities for diverse social interactions due to higher population density and proximity to urban amenities, whereas suburban areas necessitate longer travel due to lower land-use density and limited local opportunities. The study contributes to the growing literature on activity-space segregation by identifying a specific condition (i.e., travel distance threshold) required for meaningful social exposure.