Mismatch of social inclusion and street space quality: characteristics, mechanisms, and implications

Achieving high social inclusion and high-quality built environments is central to sustainable urban development. We examine the mismatch between street space quality (SQ) and social inclusion (SI) in Beijing by mapping their spatial patterns and identifying correlates of mismatch transitions. Using 14,901 1-km grids, we quantify SI across sociodemographic dimensions using the Shannon diversity index with entropy weights. We assess SQ at ∼250,000 locations from nearly one million street view images using a streetscape-based evaluation framework. SI exhibits a clear core–periphery structure, whereas SQ is highly heterogeneous and weakly aligned with SI. Mismatch maps show SI > SQ concentrated in central districts and SI < SQ concentrated in peripheral areas. We frame SQ as spatial supply and SI as social demand, and test a friction-driven filtering mechanism for mismatch transitions. Mechanism analyses further show that the two opposite transitions—toward Quality Lag and toward Inclusion Lag—exhibit largely mirrored linear associations. Specifically, NTL, LUM, RND, DEM, and PM2.5 are significantly associated with a higher probability of transitioning from Matching to Quality Lag, while showing significant negative associations with the transition to Inclusion Lag; DSS and NDVI exhibit the opposite direction. Restricted cubic spline models further reveal pronounced range dependence, including threshold-type switching, diminishing marginal effects, and local irregularities—such as tail non-significance or transient non-monotonic segments—indicating that marginal associations are not uniform across the covariate distributions. Together, the results motivate targeted, scenario- and range-sensitive interventions to curb excessive mismatch and support inclusive, sustainable urban development.