BACKGROUND: China's low male statutory retirement age (60) and its rapidly aging population make the causal effect of retirement on healthcare expenditures a pressing policy question. Centralized drug procurement, nationwide DRG/DIP payment reform, and expanded insurance coverage have reshaped the cost environment in which retirement‑related spending decisions occur, calling for up‑to‑date causal estimates. METHODS: We use the most recent publicly available wave of the China Health and Retirement Longitudinal Study (CHARLS 2020, Wave 5) and apply a fuzzy regression discontinuity design that exploits the male statutory retirement age as an exogenous source of variation. The sample consists of urban males aged 55-65 (N = 2,538), the group for whom the retirement cut‑off is unambiguous and whose healthcare‑expenditure decisions are most directly affected by formal urban insurance systems. Expenditures are disaggregated into self‑treatment (medications and preventive products) and institution‑related (outpatient and inpatient) categories, and comprehensive mechanism tests are conducted across spending ability, health conditions, time availability, and lifestyle channels. RESULTS: Retirement significantly increases self‑treatment expenditures, with out‑of‑pocket self‑treatment rising by approximately ¥14-16 per week at the age‑60 cutoff. The magnitude is modest, consistent with the post‑2018 drug‑price reductions that exceeded 50%. Institution‑related expenditures decline; out‑of‑pocket inpatient spending falls by roughly 40% relative to the pre‑retirement baseline. The gap between total and out‑of‑pocket expenditures is narrow, reflecting the decline in the national out‑of‑pocket share to 27%. The spending‑ability channel is weak, while health conditions, physical activity, and social engagement remain the dominant mechanisms. Education‑level and household‑size gradients are present but modest in magnitude. CONCLUSIONS: The retirement‑expenditure nexus in urban China is characterized by a shift toward self‑treatment and away from institutional utilization. Post‑2018 reforms have shaped the economic magnitudes of these effects, producing a contemporary benchmark for retirement‑policy calibration and health‑insurance design. Extensions to female and rural populations are warranted. CLINICAL TRIAL NUMBER: Not applicable.
Wu et al. (Sat,) studied this question.