Health Shocks, Perceived Mortality Risk, and Financial Preferences: Longitudinal Evidence From Share

Major health events are salient mortality cues and may reshape financial preferences in later life. However, evidence on the role of subjective mortality risk in shaping financial preferences is scarce. This study uses a longitudinal quasi-experimental design to examine whether incident heart attack, stroke, cancer, diabetes, or high blood pressure are followed by a change in perceived mortality risk and, subsequently, stated willingness to take financial risk and self-reported financial planning horizons. We use longitudinal data from Waves 4-9 of the Survey of Health, Ageing and Retirement in Europe (SHARE), comprising more than 200,000 observations on individuals aged 50 and older. A propensity score matching difference-in-differences approach estimates the impact of each health shock on changes in subjective 1-year mortality risk. Mediation analysis decomposes total effects of a health shock on financial preferences into components operating through perceived mortality and components acting directly. Major health events, particularly cancer, stroke, and heart attack, produce clear increases in perceived mortality risk. Changes in financial preferences are generally modest: higher perceived mortality is associated with slightly shorter self-reported financial planning horizons and lower stated willingness to take financial risk. Significant direct effects of health shocks on self-reported planning horizons remain after conditioning on perceived mortality, consistent with the influence of the background-risk channel. Overall, perceived mortality risk is an empirically relevant but only partial mediator between health shocks and financial preferences in older age. These findings underscore the importance of distinguishing mortality beliefs from broader consequences of illness when modeling household responses to health shocks and when designing interventions for individuals experiencing health-related uncertainty.