Friday, December 04, 2009


No, not for me. But Harvard has announced its plans to encourage faculty to retire. I won't call it an "early retirement" package, since it is for people over 65. Though I suppose that is early for academia.

Note that (according to the article) Harvard's Faculty of Arts and Sciences will have 127 offers to its 720 (junior and senior) faculty. So I conclude (as of next year) 1/6 of Harvard's faculty would be over 65. And the article states that the average age of tenured Harvard professors is 56. Can people from elsewhere tell me if this is unusual? Harvard has a reputation for having an older faculty; this seems to confirm it. Does his suggest a lack of planning somewhere along the line? Or is this a good thing?

I don't expect drastic changes arising from this plan; it will be interesting to see how many faculty take the offer. In general, however, is it a good idea for a university to "encourage" retirement for older faculty? And if so, what means should they use to do it?

Viewed as an optimization problem, one can ask what is the "best" distribution of faculty ages at a university, and what mechanisms could (and should) be used to maintain that distribution?


Doc Terror said...

There are two concepts behind planning at the University: suboptimization and "Apres Moi Le Deluge"

At any university you find a bunch of departments who are headed by people who are good professors, not good managers. Then there's an effectively infinite gap between them and the top administration. The top administration is continually making problems for the departments, but it's never there to help, other than in the sense of occasionally unleashing a shower of money to support the fad of the decade... You know, people caught on that "Materials Science" overpromises and undelivers, so they need to pretend they've got something new called "Nanotechnology."

The fact is that the Bill Gates and the Sam Waltons in the world have put you in a zoo that submits you to very unusual pressures that shape you into the kind of person can can impart a false consciousness at your students at a critical age. Frankly, Harvard doesn't give a damn about the rest of us so why should we give a damn about Harvard?

Daniel Lemire said...

In Quebec (Canada), men (all men!) are 50 years old on average. Full professors are 55 years old on average (according to the FQPPU).

Certainly, many of my colleagues are over 55.

Geoff Knauth said...

Albus Dumbledore made it to 116, managed his own exit, and helped defeat the Dark Lord. When it is time to go, go in style, and not too soon.

Yuriy said...

MIT's been doing this for a few years as well.

Paul Beame said...

For many years we have had a policy that is similar in intent and seems simpler. Upon retirement, faculty can start to draw their TIAA-CREF retirement funds but they also have the right to continue on for up to five years at 40% of their salary, provided they continue to do some teaching (typically one course) if funded from central university funds (or research if funded from other sources). After that it is possible but unusual (and entirely at the department's discretion) to continue at 40%. One other item that makes the continuation less desirable over time is that retired faculty are not eligible for merit salary increases.

Andrew Eckford said...

Say the average age for achieving tenure is 36. Say the average professor doesn't leave until 70. (Both these figures seem plausible considering my own institution.) Then the steady-state average age of a tenured professor should be 53. Harvard seems a little old but not outlandish.

Michael Mitzenmacher said...

Andrew --

Is it really that the average faculty member doesn't retire until age 70? I guess that's possible these days given improvements in life expectancy, but it's still somewhat surprising to me.

I suppose one thought is that as life expectancy increases, more incentives may be required to encourage people to retire, if encouraging retirements (for whatever reason) is a goal.

Adam B. said...

I have no evidence for this other than anecdotes, but I have heard that that top schools tend to hire faculty later in their careers. The anecdote I recall is a Harvard professor telling their graduating PhD: "You'll teach here some day, but you won't get your tenure here."

In your experience, do you feel more faculty are hired after producing significant results? This is probably more common in humanities than CS.

Michael Mitzenmacher said...

Adam B.

I can't really comment on the rest of the University, in that I don't have good data (not having ever checked it out). Certainly I've heard people (especially junior faculty) in other departments with anecdotes similar to yours.

In computer science, we have, I think actively been trying to hire strong junior people, and have them get tenure. We do still hire people at the senior level, but it seems pretty rare (Greg Morrisett being the last example). And not everyone gets tenure. But we've tenured several people "internally" since I've been at Harvard -- David Brooks, David Parkes, Salil Vadhan, and me -- and I'm optimistic we'll have more examples soon enough.

I'm pretty sure in CS our average age is now well below the Harvard average.

Greg Morrisett said...

[From a Virgin flight to San Fran -- thanks for the free Wifi Google.]

Yes, I've seen comparisons with other Universities (and schools) and FAS at Harvard is relatively "old" (but not by much.) For example, the B-school has a lower average age, and I suspect there incentive program has something to do with it.

By the way, the primary purpose of this offering is to make it easier to plan. As it stands, departments request to replace someone who is "going to retire any time now", but no one knows when it will really happen. In part, this is because a chair has to be very careful bringing up the issue to avoid any hint of age discrimination, so most of them just avoid the topic all together.