An alternative perspective on the relationship between multiscale built environments and walking behavior: evidence from the government-subsidized resettlement towns in China

Disadvantaged population subgroups have been increasingly displaced to low-investment neighborhoods at the urban periphery, contributing to concerns about public health and inequality in healthy living environments. This study focuses on Chinese large-scale government-subsidized resettlement towns and examines what significant neighborhood-level mesoscale and street-level microscale environmental factors are and to what extent they affect residents’ (n = 512) socio-environmental perceptions and walking behavior. Sociodemographic characteristics, socio-environmental perceptions, and walking behavior were measured in an online survey, followed by focus-group interviews. Built environment features were assessed from GIS data and audited by urban designers. Using Structural Equation Models (SEM), this study shows that neighborhood-level mesoscale built-environment factors are more associated with walking behavior than street-level microscale features. Moreover, socio-environmental perception plays a full mediating role in the association between objective built-environment factors and walking behavior but plays a partial role in the association between sociodemographic factors and walking behavior. Our findings indicate that neighborhood-level characteristics (housing type mix, morphology, and density) had greater influences on walking behavior in disadvantaged resettlement neighborhoods. To promote walking behavior in low-income communities, policy-makers and designers should also consider supporting older and wealthier relocated groups in mingling with the relatively young and low-income newcomers.