Scholarlink

<p>Understanding the determinants of walking to school is essential for promoting equitable and sustainable mobility. In highly unequal cities of the Global South, such as São Paulo, walking is often not a choice but a necessity, especially for students in low-income areas. This paper investigates how socio-spatial inequalities shape walking behavior in school commutes and challenges traditional assumptions from the walkability literature. We use data from São Paulo’s 2007 and 2017 origin–destination surveys, combined with open geospatial datasets, to model travel mode choice for public and private school students. A Random Forest model is employed to capture complex relationships between built environment factors, student socioeconomics, and trip attributes. Shapley Additive Explanations (SHAP) is applied to interpret variable influences and compare results with traditional logit models. The findings reveal that sidewalk and infrastructure conditions are significantly worse in peripheral areas, where most low-income and non-white students live. While the built environment has limited influence on walking behavior in the model, income, car ownership, and school type emerge as dominant predictors. This contrasts with the literature from more equitable contexts, where walkability infrastructure often predicts walking rates. Our results support the notion of “compulsory pedestrians” and emphasize the role of structural inequality in shaping mobility patterns. We argue that applying machine learning offers new insights into these relationships, especially when paired with explainability tools like SHAP. For cities in the Global South, walkability improvements must be targeted and equity-driven, rather than assuming infrastructure alone can induce behavioral shifts.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Assessing walking to school inequalities in a Latin American city

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月10日

查看全文
  • Bruna Pizzol,
  • Mariana Giannotti

摘要

Understanding the determinants of walking to school is essential for promoting equitable and sustainable mobility. In highly unequal cities of the Global South, such as São Paulo, walking is often not a choice but a necessity, especially for students in low-income areas. This paper investigates how socio-spatial inequalities shape walking behavior in school commutes and challenges traditional assumptions from the walkability literature. We use data from São Paulo’s 2007 and 2017 origin–destination surveys, combined with open geospatial datasets, to model travel mode choice for public and private school students. A Random Forest model is employed to capture complex relationships between built environment factors, student socioeconomics, and trip attributes. Shapley Additive Explanations (SHAP) is applied to interpret variable influences and compare results with traditional logit models. The findings reveal that sidewalk and infrastructure conditions are significantly worse in peripheral areas, where most low-income and non-white students live. While the built environment has limited influence on walking behavior in the model, income, car ownership, and school type emerge as dominant predictors. This contrasts with the literature from more equitable contexts, where walkability infrastructure often predicts walking rates. Our results support the notion of “compulsory pedestrians” and emphasize the role of structural inequality in shaping mobility patterns. We argue that applying machine learning offers new insights into these relationships, especially when paired with explainability tools like SHAP. For cities in the Global South, walkability improvements must be targeted and equity-driven, rather than assuming infrastructure alone can induce behavioral shifts.