VIRTUAL Lisbon Livable Cities

Abstract

The built and social environments within urban settings play an important role in shaping residents’ health and well-being. Livability indices bring together data from multiple domains across these environments to construct a composite measure of urban livability, serving as valuable tools for setting urban planning and public health priorities. However, most published indices lack spatial granularity, focusing on large geographies like cities or counties. This study introduces a novel small-area index for urban livability using publicly available data for 2,101 census tracts in New York City. The index includes 38 variables across six domains: socio-economic, housing, resources, environment, safety, and accessibility & transportation. Principal components analysis (PCA) was applied in two steps to construct a livability index: first to construct domain sub-scores by obtaining weights from factor loadings for individual variables within domains; and second to construct an overall index by obtaining weights from factor loadings for domain sub-scores. All variables were log-transformed and normalized prior to PCA. The validity of the index was examined through correlations with health outcomes. Domain contributions to the overall index were: resources (19.2%), accessibility & transportation (18.1%), environment (17.9%), housing (16.1%), socio-economic (16.0%), and safety (12.7%). A higher livability score was associated with lower prevalence of poor mental health (r=-0.42), poor physical health (r=-0.42), obesity (r=-0.43), and asthma (r=-0.35), among other health indicators. This comprehensive analysis, leveraging small-area data, provides localized insights, highlighting opportunities for targeted interventions to improve urban livability and health outcomes within specific social and built environment domains.