Work, Family and Retirement Timing: A Relational Retirement Age Perspective

Background and purpose: An important component of retirement planning is timing. Defining “early” or “late” retirement in research is difficult, as is often depends on what is normative within one’s field, workplace or particular job. For example, the typical retirement age for a professional job (professor, engineer, lawyer, etc.) would be different from a job that requires intense physical labor. This study takes a unique approach by looking at retirement timing from a relational perspective—using the difference between one’s retirement age and the normative/usual retirement age of individuals in similar jobs/positions at their workplace as an outcome. A gendered life course lens is used to identify potential predictors of the gap between respondents’ own retirement age and what is normative in their workplace. 

Methods: Pooled data from years 2012 and 2014 of the Health and Retirement Study (HRS) was used. HRS is a national representative data that captures work, retirement, and health-related issues for U.S. adults age 50+. The analytic sample was restricted to those who self-identified as not yet retired, and plan to retire at a particular age in the future. Normative or usual retirement age was assessed by asking, “On your main job, what is the usual retirement age for people who work with you or have the same kind of job?” Respondents’ planned retirement age was their response to the question: “At what age do you plan to stop working?” The “gap” between usual and planned retirement was then collapsed into five categories: 1) five or more years earlier than usual, 2) within five years earlier than usual, 3) same as usual, 4) within five years later than usual, and 5) five or more years later than usual. Multinomial logistic regression was used to test hypotheses. 

Results: A one unit increase in work enjoyment is associated with a decrease in the relative log odds of retiring much earlier than usual retirement age (β =.42, p<.01). A one unit increase in household income is associated with an increase in the relative log odds of retiring slightly early (β=.10, p<.05). A one unit increase in education is associated with an increase in the relative log odds of retiring much later than usual retirement age (β =.12, p<.01). Having less demanding jobs (β =.31, p<.01) and work flexibility (β =.34, p<.01) are also related to later retirement. Differences for men and women were observed.

Conclusions and implication: Examining retirement timing in a relational context enables us to not assume that 65 is the normative age of retirement in all professions and workplaces and the metric by which we should define early or late retirement. By looking at work and family predictors and how they differ by gender using this more contextualized, relational approach, we gain a better understanding of factors that relate to non-normative retirement decisions. This information can be used by employers, employees, and policy makers for building more transparent and open environment for discussing retirement-related issues and developing effective retention and retirement policies.