Multilevel Resource Capacity and Inequality in Community Eldercare Access Among Older Adults in China: Household Living Arrangements as Conditional Contexts

Background/Objectives: Access to community eldercare reflects the joint distribution of economic resources, social protection, local infrastructure, and household circumstances. Drawing on the capability approach, this study examined whether multilevel resource capacity was associated with a survey-derived access-opportunity composite among older adults in China and whether household size and generational composition conditioned that association. Methods: A cross-sectional analysis of the 2018 China Longitudinal Ageing Social Survey included 8578 adults aged 60 years or older. Partial least squares structural equation modelling estimated formative resource-capacity and access-opportunity composites. The primary model used documented Tukey-fence treatment and mean replacement; casewise deletion, 99th-percentile winsorisation, exclusion of service use, and separate-interaction specifications were examined as sensitivity analyses. Results: Resource capacity was positively associated with access (β = 0.206, p < 0.001), with community resources contributing the largest outer weight (0.659). In the joint-interaction model, the association increased with household size (βint = 0.056, p = 0.003) and was stronger in single-generation households (βint = 0.088, p = 0.016). The resource-capacity association persisted across all sensitivity models, whereas neither interaction remained significant when estimated separately under lower-collinearity specifications. The primary model explained 15.7% of access variance (R2 = 0.157; Q2 = 0.012). Conclusions: Resource capacity and urban–rural location mark important inequalities in reported community eldercare opportunity. Household moderation is best treated as exploratory because it was specification-sensitive. Social work responses include benefits advocacy, service navigation, family coordination, and community-level action on territorial service gaps.