Characterization and determinants of healthy housing index in South Asian countries with application of multinomial logistic regression analysis

Abstract

Purpose The purpose of the study is to evaluate housing quality and its key determinants across selected south Asian countries using composite healthy index. Index validation and robustness: the authors assessed sampling adequacy and factorability using the Kaiser–Meyer–Olkin (KMO) statistic and Bartlett’s test of sphericity. Internal consistency was examined with Cronbach’s alpha. The authors verified stability of the factor structure via sensitivity checks using alternative cut-points and reweighting of components. Model robustness was assessed through country fixed effects, influence diagnostics and alternative specifications. Full codebook and construction steps are available upon request. To construct a comparable healthy housing index (HHI) for South Asian countries and analyze its determinants using multinomial logistic regression, providing market – and policy – relevant evidence for housing finance and urban development. Design/methodology/approach The authors use nationally representative DHS microdata from six South Asian countries. A composite HHI is developed via factor analysis with KMO/Bartlett tests and internal consistency checks. Determinants of HHI tiers (lower, middle and upper) are estimated using multinomial logit, reporting relative risk ratios (RRRs) with 95% confidence intervals and marginal effects. Robustness includes alternative cut–points, re–weighting, influence checks and country fixed effects. Findings Urban residence, electricity access, private sanitation, kitchen availability and cooling present systems are consistently associated with healthier housing tiers, whereas shared toilets (Pakistan) and pet ownership in several countries are linked to lower tiers. The results are robust across specifications and align with finance–relevant risk screening. Research limitations/implications The analysis uses cross-sectional DHS microdata, so causal inference is limited. Survey years differ across countries, and amenity variables may contain reporting or classification error. The index relies on factor-based weights and threshold choices that, despite sensitivity checks, may not capture unobserved quality dimensions. Multinomial logit assumes independence of irrelevant alternatives; violations could bias RRRs. Country heterogeneity and omitted local policies may affect generalizability. Future research should use panel/longitudinal data, causal designs, spatial dependence checks, alternative weighting schemes (e.g. PCA/entropy) and external validation against health outcomes. Practical implications The HHI offers a transparent tool for lenders, developers and municipal agencies to target upgrades, calibrate subsidies and screen credit or investment risks in rapidly urbanizing South Asian contexts. Social implications Improved targeting of infrastructure and services can accelerate progress toward resilient, equitable and health–promoting housing consistent with SDG goals. Originality/value This study integrates a validated multi country housing quality index with multinomial risk classification at scale in South Asia, bridging housing market analytics, finance and policy.