Spatial mismatch and mobility barriers: A spatially matched case-control study of special education (SPED) school students

In high-density urban structures, particularly in large cities of the Global South, the acute spatial mismatch between centralized special education resources and dispersed residential populations imposes severe structural constraints on accessibility. While transportation accessibility is a key determinant of inclusive education, the mechanisms of exclusion affecting children with disabilities in this context remain understudied. This study leverages a unique dataset of 491 children with disabilities attending special education (SPED) schools and 982 spatially matched peers in Beijing, employing a multi-stage Spatial Propensity Score Matching (S-PSM) and Firth-penalized multinomial logit (MNL) framework. We identify three interconnected mechanisms of exclusion. First, distance attenuation patterns differ markedly across student groups: students with intellectual disabilities experience a pronounced “mobility cliff” at approximately 2 km when walking, whereas those with hearing impairments can endure longer walking distances for “survival commuting.” Second, comparative analysis reveals a non-linear “inverted-U” pattern of car usage, indicating that the private coping strategy will collapse under hyper-dense urban congestion. Third, the household strategy of “renting for education” exposes the privatization of public responsibility, where families are forced to convert transport barriers into severe housing burdens. Our findings demonstrate that transport inaccessibility acts as a core mechanism of social exclusion, indicating that effective intervention must address deep-seated spatial inequities rather than transport demand alone in rapidly urbanizing contexts. In this context, we propose three policy recommendations addressing housing support, dedicated school transportation, and inclusive infrastructure.