Connecting built form diversity and housing mix with refining diversity measurement under the general framework

Creating socially diverse neighborhoods through housing mix is an essential planning approach in big cities, but the role of built form diversity in facilitating social mix remains underexplored. Although diversity is prevalent in urban design and urban form studies, its ambiguous articulation and underdeveloped measurement hamper effective assessment of the presumed relation between physical and social diversity. This study refines the measurement of built form diversity and examines its relations with housing mix at 306 neighborhoods of Amsterdam, a superdiverse city. We employ the general framework of diversity, and integrate measures to incorporate disparity, variety and balance. Specifically, categorization through a clustering algorithm based on four building attributes effectively captures the disparity of built form, supplemented by a disparity index for each attribute. Absolute and relative indices accommodate different standards of balance. Regression models yield significantly positive associations between built form diversity and housing mix, which reinforce the potential of planning diverse built form in fostering housing mix. Disparity of openness (representing street width and setback) and footprint area are positively associated with tenure or dwelling mix. These findings support the theoretical connection between physical and social diversity. The study demonstrates that measuring diversity should integrate multiple measures and dialectically handle the uncertainty of disparity and balance. Introducing the general framework provides a common ground to study and discuss diversity from different planning specializations. • Uses the general framework to characterize social and physical diversity • Integrates existing measures to incorporate disparity, variety and balance • Built form diversity is positively associated with housing tenure and dwelling mix. • Disparity of building setback and footprint area are associated with housing mix.