To Engage or Not to Engage? Social Activities and Survival Beliefs in Older Adults

While survival belief is central to major life decisions, its social determinants remain understudied. Social engagement is proven to influence various health outcomes, yet the direction of these relationships stays controversial, with quasi-experimental designs being laid aside. It is also unexplored what happens once individuals withdraw from established social life. To address these gaps, this paper leverages the potential randomness in the timing of individual (dis)engagement behavior to examine the causal effects of social (dis)engagement on the subjective probability of unsuccessful survival (SPUS) of young-old Chinese, using a panel of 11,412 respondents from the 2011–2020 China Health and Retirement Longitudinal Study. A dynamic difference-in-differences approach, with various constructs of treatment and control groups, was employed to validate the estimation. Several robustness checks, including factor analysis and propensity score matching, were further conducted. The primary findings are that social engagement reduces older people's SPUS by up to 2.80 percentage points immediately and by 6.67 percentage points one period later. Conversely, social disengagement shows no immediate effects, but its post-treatment impact can reach up to 14.76 percentage points among subgroups. Interestingly, the engagement effects persist only among non-switchers, fading once switching begins, and the disengagement effects also emerge only among non-switchers. In sum, this study provides partial evidence of stronger disengagement effects, though these effects occur neither instantly nor universally.