When to Retire?

Several years ago, the gubmnt of Ontario eliminated mandatory retirement at age 65. The result has been that professors at my university can continue to work beyond the age of 65, drawing their usual salaries, with only minor reductions in some benefits beyond the age of 65 (less life insurance and no additional university contributions to the defined-contributions pension plan).



The reaction by professors has varied considerably. For example, I have publicly proclaimed that I intend to teach until I am 90.



But I have observed, in limited conversations, that people with children generally expect to work longer than those without children. Perhaps one reason is that we have more debts or smaller net savings because we have spent more on our children and grandchildren. But another explanation offered by one (childless) person is, "I want to spend my retirement savings.... what else am I going to do with it?" which is another way of saying that those of us who plan to work longer have in mind, as one of our objectives, leaving some of our wealth to our children.



This is all ad hoc empiricism, of course. But my hypothesis is that people who have children and grandchildren, because they have interdependent utility functions (with the utilities of their progeny entering with a positive derivative!), will in general tend to work longer than those with no children or grandchildren. They will typically have given more to their offspring and hence want to work longer to have more during their later years and/or they may want to reduce the risk of running low on funds late in life, knowing that anything left over when they die will go to their children and grandchildren.



Those without children may well have other persons or groups to whom they have given or wish to leave some wealth, but my suspicion is that these altruistic sources of utility are, typically and on average, less significant than having offspring; i.e., I am hypothesizing that charities and non-relatives enter most people's utility functions with smaller weights than do our children and grandchildren. But these speculations are all a priori, based on casual empiricism. Does anyone know the literature on this topic?