A forward-looking age based on longevity expectations

Arda Aktaş, of Stony Brook University, USA, investigated how “subjective age”—how long people think they have left to live—changes with characteristics such as gender and education.

Arda Aktaş

Arda Aktaş


Many personal decisions are shaped by people's expectations of the future. However, such expectations are rarely included in the study of those personal decisions. Often, studies that analyze forward-looking decisions use chronological age, an inherently backward-looking measure, as a proxy for those expectations. Implicit in this approach is the notion that all groups move through life-course stages in a chronological lockstep. Quite to the contrary, however, we can actually observe that different groups behave differently even though they may all be members of the same birth cohort. There are many reasons for this heterogeneity, including the fact that perceptions of aging may not be the same for all individuals because they have different characteristics. Thus, depending on which particular stage individuals are in at a given point of time, their behaviors will be different from those of other members of the same cohort with different characteristics. However, while the perceptions of aging cannot be directly observed, an individual's perception of aging can be captured by linking it with subjective life expectancy, that is, how many years an individual thinks that she/he has to live. Subjective life expectancies are generally obtained in the form of survival beliefs, that is, the probability of surviving up to a specified target age.

We propose a method to quantify people's longevity expectations using subjective survival probabilities and also the technique of [1] to transform it to an index measured in years. This will make it easier to use in any analysis where people's expectations matter and also make it comparable with the conventional age measure. We call this new approach of measuring age “forward-looking age.” This alternative age measure can contribute to existing literature by providing new insights into the examination of individual decision making.


We use a two-part methodology to compute a forward-looking age that is based on data of longevity expectations collected in the Health and Retirement Study. In the first part, we propose a method to translate those expectations into life tables. We tackle the focal points problem using a random effects ordered probit model to obtain refined probabilities that depend on the characteristics of each individual; we then use these refined probabilities to conduct an NLLS estimation of subjective survival functions and to construct life tables for groups with various characteristics. In the second part, the life tables are used to produce forward-looking ages that can be used in the study of forward-looking decisions.

Results and conclusion

We find that there is substantial variation in the forward-looking ages of individuals with different characteristics (such as gender, cohort, education, place of birth, adverse health conditions, and smoking habits) and that this variation tends to increase with chronological age. In particular, we observe that education matters for both genders, but the magnitude of its effect is larger for women. Moreover, the presence of any particular health condition or smoking habits increases the forward-looking age. Therefore, the effect of smoking or having any adverse health condition is larger for lower-educated groups compared to higher-educated groups. Finally, the effect of education is higher for women of the younger cohorts. For men, there is no significant difference in terms of education among cohorts.


[1] Sanderson WC, Scherbov S (2013). The characteristics approach to the measurement of population aging. Population and Development Review 39 (4), 673-685.


Warren Sanderson, World Population Program, IIASA


Arda Aktaş, of Stony Brook University, USA, is a citizen of Turkey. She was funded by the IIASA US National Member Organization and worked in the World Population Program during the YSSP.

Please note these Proceedings have received limited or no review from supervisors and IIASA program directors, and the views and results expressed therein do not necessarily represent IIASA, its National Member Organizations, or other organizations supporting the work.

Print this page

Last edited: 02 February 2016

International Institute for Applied Systems Analysis (IIASA)
Schlossplatz 1, A-2361 Laxenburg, Austria
Phone: (+43 2236) 807 0 Fax:(+43 2236) 71 313