|| Checking for direct PDF access through Ovid
Bereavement is an inevitable event in our life. This paper employs the Taiwanese panel Survey of Health and Living Status of the Elderly (SHLSE) to evaluate the impact of losing a spouse on self-assessed health and subjective well-being measured by depression and life satisfaction. Propensity score matching methods are used to generate a hypothetical bereavement date and a weight for the non-bereaved to create a comparable non-bereaved cohort and a difference-in-differences (DiD) approach is used to estimate the impact of spousal bereavement.The results show that spousal bereavement increases depression scale by 1.81 points but this increment decreases by 0.43 points every year after bereavement. It takes approximate 4 years to restore to the level prior to bereavement. We also examine the demographic and socioeconomic differences in the spousal bereavement impact and find that the spousal bereavement impact is greater on the bereaved in the higher income group in terms of self-assessed health and depression. Our results only represent a lower boundary of the possible impact of spousal bereavement on self-assessed health and subjective wellbeing due to data restrictions.