Retirement and Interdependent Utility Functions

A few days ago, I was speaking with a colleague who is nearing retirement. I told him I am planning to teach until I'm 90.

His reaction was,

"Why? We have a good pension plan (defined contribution), and you can't take it with you. I want to retire and spend it."

He has no children and he expects his wife to die about the same time he does.

Contrast his decisions with those of someone who has a spouse who is expected to live a long time and who has children and grandchildren, to whom s/he would like to leave a sizeable financial legacy (or on whom s/he would prefer not to become dependent, financially and otherwise). The latter person is much more likely to work longer.

I don't know the literature in this area, but I would predict that ceteris paribus, people without children and grandchildren retire earlier than people with children. Here is one study with results supporting this prediction, though my argument depends heavily on the assumption of interdependent utility functions [I get utility from my children's and grandchildren's happiness and would like to leave them some money after I die] and depends less on whether the children are under 19 years of age.

The concept is the same one I wrote about yesterday:

The answer quite clearly is that I care about my children, my grandchildren, and others; I have an interdependent utility function. The utility of these people enters into my utility function in such a way that I am willing to trade-off some of my utility over the next ten or so years in exchange for improving their chances of avoiding an asteroid collision.

... There is probably an analogy to caring about what happens to the Social Security programme in the U.S. If you think it is doomed and disaster will ensue in 30 years, you might not care if you're an executive with AARP/CARP, but you are more likely to care if you worry about the lives of your children, grandchildren, and others.

Topic: Economics