Work-to-Retirement Transition and Physical Activity Change: Causal Evidence From China.

Abstract

Population aging and increasing retirement rates are reshaping physical activity (PA) and health among older adults worldwide. Using the China General Social Survey data (CGSS), this study applies a fuzzy regression discontinuity design to estimate the causal impact of retirement on PA among Chinese older adults. We find that retirement significantly increases PA frequency. The effect is amplified by higher income and better health, but is weakened by psychological stress. Causal forest analysis reveals pronounced heterogeneity: income consistently emerges as the most influential moderator, while the importance of social participation, intergenerational caregiving, and urban-rural residency rises notably once provincial clustering is considered. Regional disparities also persist, with stronger effects in the eastern provinces, contrasted by more mixed and uneven patterns in central and western China. These findings highlight heterogeneous aging trajectories, showing that retirement's impact on PA is context-specific and requires locally adaptive approaches.