How do subjective life expectancies compare with mortality tables? Similarities and differences in three national samples

Estimates of personal longevity play a vital role in decisions relating to asset accumulation and decumulation. Subjective life expectancy (SLE) is a measure of individuals’ expectation of remaining years of life. Either explicitly or implicitly, it is a key determinant of consumption and savings behaviour, and may be guided by a person’s own health and health behaviours. The Gateway to Global Aging, a platform for the Health and Retirement Study’s (HRS) family of population surveys, provides harmonised longitudinal datasets for many countries, each based on individual survey responses from respondents aged 50 and above. In this paper, we analyse SLE three of these datasets: the English Longitudinal Study of Ageing (ELSA), The Irish Longitudinal Study of Ageing (TILDA) and Healthy Ageing in Scotland (HAGIS). First, we focus on measurement of SLE, followed by the SLE differential – the discrepancy between SLE and mortality risk indicated by population life tables. One novel finding from our analysis is that the SLE differential is positive for Ireland and is negative for Scotland and England. This difference does not appear to be explained by differences of survey design or population characteristics.