'If we were disaccredited, it would be the accrediting agency that looked foolish.' Harvard's departing president Neil Rudenstine tells Auriol Stevens about the challenges of running a world-class university.

Neil Rudenstine leaves Harvard University this summer after ten years as its president. During his term of office he has become famous mainly for raising $2.6 billion (?1.8 billion), more even than the university's ambitious target. He has said little about ÖйúAƬ in general or about the challenge of managing Harvard's grand barons in particular. There have been bumps along the way, inevitable perhaps in a university famous for its decentralised structure, with "every tub standing on its own bottom". Each faculty is responsible for its own programmes and its own fundraising, meaning Harvard's president has limited formal power.

Yet Rudenstine has played a pivotal role. Before the university-wide appeal for money was launched in May 1994, Rudenstine insisted on an 18-month planning exercise to set priorities and guidelines and ensure the university presented a coherent message to the outside world. Some say the resulting restructuring, including five major inter-faculty initiatives, only one of which - on environment - has taken off, was more apparent than real, with the major schools quickly reasserting their autonomy. But it served its purpose. The fundraising campaign rolled forward and endowment income now covers 28 per cent of Harvard's costs, compared with 18 per cent in 1991.

"Yes, we were successful. Of all the goals, we probably fell short on just two. One was financial aid - fellowship aid for graduates and professional school students. The second was that we had hoped to endow more faculty chairs, especially junior positions," Rudenstine says. Recent stock market falls have had some impact on endowment income, but it has fallen far less than the market generally. "We have skilful managers who saw what was coming, got out of some things and got into others."

ÖйúAƬ

One of the things they got into was real estate in the Boston area, causing muttering among the locals because charities' exemption from property taxes puts more load on other residents. In 1999, the university sought to demonstrate its civic virtue by launching a $20 million, 20-year low-interest loan fund for affordable housing in the Boston/Cambridge area, called Harvard 20/20/2000. Among guidelines for the campaign were some covering the role of donors. "We have walked away from some large gifts because we just didn't think they were in line with our priorities, or we thought that the donor would want controls that we didn't think a donor should have, such as wanting to designate an architect for a building, or wanting to design a programme with restrictions that would not be appropriate or would not be likely to stand the test of time."

Donors may be involved on advisory committees, he says, "but we would never have a donor in a position of power and we would certainly never have a donor on a selection board".

ÖйúAƬ

Rudenstine believes private money is essential for maintaining quality. Having visited more than a dozen countries during his presidency and talked to many people in ÖйúAƬ around the world, he believes that even universities in rich countries are in trouble. "To the extent to which they depend on the government for all or most of their money, they are looking at a comparatively bleak future," he says. "These days, and for the foreseeable future, you will never get ÖйúAƬ high enough on the priority list."

The result, he says, is that you have hordes of students and very few faculty members. The problem of access has been addressed, but not the problem of retention. "One good thing the United States did between 1950 and 1975 was to address the problem of the scale of the institution and adequate faculty resources as well as access. Trying to do the retention thing later and doing it all on government money is a non-starter."

Rudenstine's scepticism of government extends also to quality assurance. At Harvard, this rests on four legs: internal visiting committees that report to the university's board of overseers; national accrediting agencies; government-sponsored peer-review systems for distributing federal research money; and student evaluations of teaching. Internal review, organised by the board of overseers, is the most important. The 30-strong board, made up of academics and other experts, is the main custodian of quality. The board puts together visiting committees for every school and every department, drawing on academics from other institutions. These committees look at everything - research, graduate and undergraduate teaching, science equipment, libraries and forward plans - and talk to everyone, staff and students. They keep in touch with departments regularly and visit formally about every two years, reporting to the dean and the overseers, with a follow-up some three months later. "The fact that faculty know there is an overseers' committee and that there is a visiting committee gets them up to scratch, keeps them planning and keeps them trying," Rudenstine says.

National accrediting agencies matter less. "At Harvard, if we were disaccredited, we'd go on being Harvard and it would be the accrediting agency that would look foolish."

ÖйúAƬ

Nonetheless, accrediting teams are made up of distinguished people who know about ÖйúAƬ, headed by the president of a comparable university. "It is a private group from ÖйúAƬ, not a government group. These people know about education and, on the whole, governments do not. Governments pay the bills, governments can be more or less educated about education, they can understand parts of it, but there are vast parts they don't understand."

Accrediting agency reports can be constructive and useful, but Rudenstine says that "in the professional schools, they tend to be rather tunnel-visioned. You know, 'You should have more courses in structural engineering.' Well, we are not turning out that kind of engineer. We are turning out a sort of engineer who would be a leader in a broader field. We have some debate. But in the end if they don't accredit us, it's their loss."

Only one part of the quality system stems from government - the distribution of federal research funds through agencies such as the National Institutes of Health.

Individual researchers apply for and compete nationally for the grants. They are not allocated to institutions "by block", which Rudenstine considers of major importance. "That would be death. That would be horrendous," he says, adding that block grants would not ensure the best researchers are supported. "How would you know the money was not going to be given to people who were not as good as other people? Mediocre research is worse than no research."

ÖйúAƬ

This does not mean, however, that only established researchers should be funded - spotting research potential is part of supporting excellence. "You have to be quite subtle and sophisticated in deciding what is mediocre and what isn't, and what in the long run might prove to be extraordinarily truthful, even though it may look irrelevant now. But you ought to be able, within a reasonable range, to identify the people who are extraordinary and the people who aren't."

The main drawback Rudenstine sees to this part of the system is that the government has been getting meaner about funding overheads, leaving the universities to find money to support grant-winning staff. This is one of the purposes for which Harvard raises private funds. To these three official mechanisms for ensuring quality must be added termly student reports on each course. "If year after year, term after term, you get ratings that say you are a well-informed but extremely dull lecturer, everyone knows it. The feedback loops are enormous and powerful, especially at Harvard. You are in a fishbowl."

ÖйúAƬ

After ten years in Harvard's hot seat, Rudenstine offers this advice on an effective quality-control system: "Government should tread very lightly, very cautiously, in these areas. At the same time, it does have to be a two-way street. For the institutions to justify the trust placed in them, they do have to adopt their own rigorous evaluation. If you are not willing to have external people from other universities come in and talk to your faculty and talk to your students and get candid replies and then write a candid report about the fact that one-third of the department is really quite sleepy and doing almost nothing and another third is fantastic and you had better do this, that and the other, then it is difficult to trust the department or the university."