To analyze the impact of multi-level social capital on the subjective well-being of rural middle-aged and elderly individuals, exploring the roles of different levels of social capital.

Utilizing data from the 2020 China Health and Retirement Longitudinal Study (CHARLS), a multi-level linear model was constructed to investigate how individual social capital and community social capital influence the subjective well-being of rural middle-aged and elderly individuals. Instrumental variables were employed to validate the causal relationship between these dimensions.

In the null model, the intra-class correlation coefficient (ICC) was 0.135, indicating community-level differences in the subjective well-being of rural middle-aged and elderly individuals. The multi-level linear model revealed that both individual social capital and community social capital positively affected the subjective well-being, with regression coefficients of 0.018 (P<0.05) and 0.001 (P<0.05), respectively. Factors such as age, marital status, self-rated health, physical disabilities, health insurance, and education level significantly influenced subjective well-being, with regression coefficients of 0.118 (P<0.001), 0.051 (P=0.022), 0.181 (P<0.001), -0.155 (P<0.001), 0.070 (P=0.054), and -0.082 (P<0.001), respectively.

The multi-level social capital plays a positive role in enhancing the subjective well-being of rural middle-aged and elderly individuals, and efforts should be made to further improve the social integration of the elderly.