Thursday, October 17, 2013

The Crimson on the Dean's Role

The Crimson’s Jared Lucky wrote a good article about the College deanship, including one money quote from me: “Just because you don’t have a lot of budgetary authority doesn’t mean you can’t influence decisions.”

That sums up for me the most important implication of the article. It cuts against the prevailing Harvard fantasy that anyone is really in charge of anything, that we are a businesslike (if not militaristic) place with crisp lines of command and control. A decade ago, the structure was shallower and less clearly defined; frankly, it mattered less. Nobody ever referred to one Harvard official as “reporting to” another the way they do today, though of course no one was under any illusion about the fact that the person who hired you could fire you. Lately we have created so many inflated titles that the community—and the people who hold the titles—too often think that if you need an X kind of problem solved, or if Harvard hasn’t made enough progress on issue X, the Vice-Wumpus for X is the person who can fix it. (For “-Wumpus,” substitute “-Dean,” “-Provost,” or “-President.” Many of the Wumpi are good people, and I mean no disrespect to them personally; it’s their number and their structure that has some downsides.) It is also thought that the lack of anyone with “X” in the title is some kind of proof that Harvard doesn’t care about X. 

But creating an “Assistant Wumpus for X” position may only make the problem worse, because all of the other parties who need to participate in owning the problem to solve it really instead of symbolically  may withdraw, thinking X is that person’s problem. They are reluctant to invade their colleagues’ turf, and the X-Wumpus may be territorial and resent initiatives in area X that he or she did not initiate.

I have a preference for vague, encompassing, ambiguous titles. I like “Bureau of Study Counsel” and “The Harvard Foundation,” which intentionally trade clarity for dignity. It’s OK if you have to explain to people what these are; that provides an opportunity to explain what they aren’t, and why they are different from what other colleges have (a mental health center, a third world house). While I was dean, I actually went so far as to drop the “X” from the titles of several deans, so that when the Associate Dean for Finance and Administration left, I replaced her with a PhD who could do finance and administration but could also be tasked with academic issues. By the time I left I had a strong ball team consisting mostly of utility infielders. We used to sit down together a couple times a week, report to each other on what was working and what wasn’t, and divide up any new problems that had arisen as seemed best, not according to previously defined job descriptions.

The Crimson story positions the deanship of the college in a much-simplified version of Harvard’s org chart today. It looks like a tangle, but doesn’t nearly do justice to the complexities and overlapping directorates. It looks very different from what it was only fifteen years ago—compare, for example, the Provost’s office today with what it was in 2007, 2002, and 1998. And it looks like the webmaster is having trouble keeping up with the additions.

Reading the article, about how much coordinating and cooperating is needed for the dean to get anything done, I asked myself whether I got anything done in the eight years from 1995-2003. It’s easy for me to think just in terms of problems that occupied a lot of my time—peer sexual assault, for example—but those problems are complex and any measure of progress would be highly controversial. If my accomplishments were measured Washington-style, by how many pieces of legislation I got passed, it would be a short list, consisting perhaps of a single act of addition by subtraction—I got rid of the rule against running a business out of your dorm room. A fine thing to have done, but it’s hard to boast about getting rid of a rule when nobody could remember the point of it.

The one thing I did that was really hard to do actually happened before I became dean: “randomizing” the Houses, getting rid of the previous system under which students were allowed to express housing preferences and thereby achieve a measure of self-segregation. Whatever argument from “diversity” one might make for randomization, the new system had the beneficial effect of making any social problem in the College everyone’s problem. I was told in pre-randomization days, for example, that if a gay student wasn’t happy where he were living, it wasn’t really a problem, since he could always transfer to Adams House, which knew how to make people like that feel comfortable. No one would say such a thing today (about the residential system, at least).

That randomization recommendation came out of the “Lewis-Maull” Committee on the Structure of Harvard College which I co-chaired. It issued its report in 1994 and, officially, my predecessor Fred Jewett was the dean who implemented it. Alas, Fred didn’t bother to implement the footnote about randomization in the Lewis-Maull report that called for controlling the gender ratio in each House. Perhaps he thought that all the Houses would come out about 50-50, when statistically it was almost certain that in at least one House the ratio would be lopsided. That House was Currier the first year, and it was not pretty; the oversight was fixed the following year.

At first the Masters were mostly, though not unanimously, in favor, and the student body was mostly opposed. (After all, it takes a lot of respect for the common good to answer “I don’t” when faced with the question “Do you or don’t you want a choice in where you are going to live?”) Maybe nothing I did subsequently as dean had as much lasting significance as the jawboning I did for a couple of years to persuade the key people, especially key student leaders but also alums and Masters, that randomization would not be the end of the world.

That was hard work of enduring importance. It is pretty amazing that it’s a non-issue today; ask students today and they will be astonished at the idea that the system could be anything other than what their predecessors considered an unthinkable abomination twenty years ago. There have been no five-year reviews or student surveys that I know of, no UC bills calling for a return to student choice. I can imagine, however, that the issue may come around again if Houses get built in Allston: will engineering students and athletes be allowed to express a preference to live near the facilities where they spend most of their time?

I wonder how the decision-making about such an issue would be handled today, with so many more Vice-Wumpus for X positions, each with an opinion about how such a change would affect an identifiable population, with so much more centralization of control generally, so much more sensitivity to messaging, branding, and communications instead of principle and substance, and with so much more confidence that student opinion polling and “best practices” of other colleges are relevant to the formation of good educational policy for Harvard. Hard as it was at the time, I wonder if it could even happen today. Randomization might have happened at the last possible moment when Harvard’s undergraduate House system could have been fixed. 


  1. "I was told in pre-randomization days, for example, that if a gay student wasn’t happy where he were living, it wasn’t really a problem, since he could always transfer to Adams House, which knew how to make people like that feel comfortable. No one would say such a thing today (about the residential system, at least)."

    Still trueish. Last year, it was the case that freshman entering the lottery had the option of requesting 'gender-neutral housing' in Adams, Cabot, Kirkland, Pfoho, or Quincy (the specific house was, of course, random), as those houses had more-relaxed mixed-sex rooming policies.

    Topically: "The pilot program comes in the wake of plans announced by [Dean Hammonds], to appoint a director of BGLTQ student life and to accept the recommendations of the BGLTQ Working Group..."

  2. Humans in general do not appreciate or trust randomness. I'm reminded of "Fooled by Randomness" by Nassim Taleb ( It's a good read. I wasn't aware that you had recommended using randomness but then correcting for gender balance. Interesting idea, but is gender the only thing to correct for?

    (We had mixed-sex bathrooms in my MIT dorm back in 1983-1987, but only mixed-sex bedrooms only informally, when mutually consenting people swapped rooms de facto but not on paper.)

    1. We discussed that question, but there is another wrinkle that may not be obvious and that makes this the right decision. The Houses cover 3 classes, sophomores, juniors, and seniors. So a third is replaced every year and the other two thirds persist. Because the numbers of each gender are large, a significant deviation from the average is noticeable immediately. Not so for the various minority groups (minority in the mathematical sense, including athletes, etc.). A one-year variation just mixes in with the remaining two classes and nobody notices, and it is almost certainly not going to recur. Even if there is a variation in the same direction two years running, there is no reason for it to persist, so no House gains a minority identity as used to happen.

    2. One other amusing point. The first year was the roughest, not just because of the lack of gender control that led to the disproportion in Currier House. The Houses that already had distinct populations because of selection effects noticed the change the most. One House that had a reputation for quiet, studious types happened to get two blocks of party animals; that would not have been a big change from an average, but it was a big change from the decidedly non-random prior distribution! Suddenly Saturday nights were noisy.