I have to initially suspect that this difference between the genders has at least something to do with the fact that often -- particularly in the eras of the currently elderly -- the "work" of a female extended well beyond the office and into the household. Retirement may end the office duties, but the household duties will continue.

I doubt that contributes, because it's not just limited to humans. Females of most species live longer than males.

he is suggesting it as an explanation for retirement contributing to premature death in men more than in women, not the difference in life expectancy in general.

It could be related to post-retirement activities.

Men, as a hypothetical, might take up drinking and fishing, leading to an increase in mortality caused by liver failure, cancer, what have you.

Women, again as a hypothetical, might take up other activities that don't increase mortality.

It is hard to determine the causal relationship; I think it goes in both directions: those who are closer to death are more likely to retire and those who retire are more likely to die sooner.

certainly we would expect that people in good physical health are less likely to retire, that those in poor health are more likely to.

But...

> Finally, we nd some evidence in line with the view that involuntary early retirement has a negative impact on health, but not necessarily voluntary early retirement.

So I guess I'll keep my early retirement plans.