There were many interesting comments, as usual. I was particularly impressed by a comment by a surgeon, who pointed out "that most surgeons as they grow older, find the demands of long, difficult cases beyond their capabilities in terms mainly of stamina and fatigue. This leads to attempting shortcuts which often lead to bad results." I think much the same thing can be said of some judges, whose shortcuts involve excessive reliance on the views of their colleagues or their law clerks. Another comment points out that we use testing heavily to determine fitness of young people for various jobs, schools, etc. We do so because we have reasons to distrust the capacity of young people; we have different but not weaker reasons to distrust the capacity of the very old, when they occupy key jobs, blocking the advancement of younger people.
What I do think I exaggerated was the zero-sum character of forced retirement. If Justice X retires earlier than he would like to and is replaced by Judge Y from a lower court, then Y has in turn to be replaced, and so there may be a net loss of output to the extent that X is still productive. But since the Supreme Court is more important than any lower court, the net gain from replacing X by Y is likely to exceed the net loss from replacing Y in the lower court by someone who on average will not be better (for Y is not old, so there is no presumption that his replacement will be better).
I disagree with the contention in one very interesting comment that "Americans tend to stow their old people away in Florida, or shelve them in old folks homes." We do have a youth culture, in the sense that the elderly are not respected as such, that is, for the wisdom of their experience; this is because the pace of change in our society is so rapid that the experience accrued by the elderly tends to obsolesce quickly. (As one commenter put it, "I see many people who are not demented at all but emotionally wedded to the ideas of the 50s and 60s, doing more harm than good in their professional capacity.") But the effect of increased longevity in a democratic system is to increase the power of the elderly, and that is everywhere evident.
Several comments suggest that there is a market failure in the employment of the elderly--that in the absence of legal protections, there would be rampant discrimination. I don't think that's likely. Employers want to minimize their quality-adjusted labor costs. An elderly person who is still productive should be able to retain his existing job or find a new one. Of course markets don't work perfectly, but neither--to say the least--does government regulation. In this connection, one commenter, a college dean, suggests that the real problem is that because salaries at his college are based on seniority rather than productivity, replacing equally productive old with young can reduce employers' costs, but is forbidden. However, there is no requirement that salaries be based on seniority (maybe some state law requires that in the particular college of which this person is dean, but that would hardly be evidence of a market failure). Federal judicial salaries, for example, are not tied to seniority at all.