The manufactured housing mosaic: an intersectional approach to measuring social vulnerability

This article presents a novel method for identifying and analysing the diverse pathways through which housing type is linked to social vulnerability. We use manufactured housing (MH) to challenge extant homogenizing, damage-centred narratives, and apply a novel method for identifying and mapping distinct housing-linked vulnerability profiles. This mixed methods approach applies logistic principal component analysis (LPCA) to census microdata for 1094 MH households in Pima County, Arizona, revealing four distinct household profiles with overlapping constellations of vulnerability drivers. These profiles are then validated using a combination of statistical methods and qualitative data. Interviews with 72 MH residents are used to provide concrete illustrations of the abstract vulnerability profiles identified by LPCA. Each profile is mapped at the census block group (CBG) level to gain insights into their respective geographies. We find that MH households, even among the vulnerable, are highly differentiated, with unique relationships to various vulnerability-linked processes, and distinct geographies and potential relationships to hazards and other risks. We conclude by elaborating possible policy implications of these findings.