Dozens of academic institutions, even small colleges, have endowments of at least $1-billion, despite losses following the burst of the dot-com bubble. Because colleges and universities see a generous endowment as a top priority, endowments have grown faster than budgets. That prompts two questions: When is an endowment big enough? And can an endowment be too big? There must be at least some rough ideal value for an endowment. Certainly endowments can’t outpace institutional budgets forever.
The issue is not of importance only to the bean counters. It affects an institution’s nature, mission, and atmosphere at every level. I teach at Grinnell College, in Iowa, a long way in miles and dollars from the astronomical endowments of Harvard, Stanford, and Princeton. Most folks figure that those big-name, big-money universities are the models that all colleges would wish to emulate. Or, if we restrict our attention to liberal-arts colleges, you might expect that places like Williams, Amherst, and Swarthmore would explore the upper reaches of the endowment stratosphere before places like Grinnell.
But I would argue that Grinnell has already reached, and passed, an ideal endowment size, and that for Grinnell and many other colleges, an overly large endowment is both risky and difficult to manage. Having spent a half-dozen years as an amateur investigator of issues involving the endowment and two years serving on Grinnell’s strategic-planning committee, I have some firsthand experience that I hope will be helpful to others.
Rather than consider Grinnell per se, and risk getting sidetracked into technicalities of accounting, let’s consider the hypothetical liberal-arts institution Mythic College, which is quite similar to Grinnell, in the year 2005. Mythic has an annual base budget of $65-million and an endowment of $1-billion. Over the past 20 years, the budget has increased at an average rate of 6 percent per year, and the endowment principal (after income and payout to the college budget) has increased at an average rate of 10 percent per year. Mythic’s board of regents has established a prudent endowment payout rate of 4.5 percent per year. Arithmetic places the endowment payout at just shy of 70 percent of the college’s operating budget.
We can logically assume that growth of both budget and endowment will continue, perhaps with a few bumps here and there. So 2006 is likely to see a $69-million budget and a $1.1-billion endowment, with 72 percent of the budget coming from the endowment payout of $49.5-million.
Continued extrapolation puts the endowment payout in excess of the budget a decade later, in 2015. That projection suggests we should consider a major change from business as usual. But endowment income is not like a salary. The stock-market woes of the past half-dozen years certainly sent that message loud and clear. What could happen at Mythic College in 2015? Say the Board of Regents eliminates tuition and fees and disbands the development office. But then in 2016, the stock market collapses, and the admission yield for the incoming class is much higher than predicted. The board demands a sudden return to tuition, with faculty and staff cutbacks that lead to bad press off the campus and a drop in morale on the campus — certainly a scenario to give trustees and administrators pause.
So if our projections can’t reliably guide us, is there a better measure we might turn to?
Endowments are usually compared using two figures: the total value and the value per student. But neither is really the correct figure to consider when attempting to find the ideal endowment size. Even two institutions of the same size may have very different costs of operation. The most logical figure of merit is the ratio of the endowment to the full operating budget, which we can call the endowment-expense ratio — a universal measure of the ability of an endowment to pay for the activities of an institution.
Although the notion of comparison by the endowment-expense ratio is not new, only recently have there appeared detailed analyses correlating institutions’ behavior with excess endowment-expense ratios. A recent example, an influential article by John E. Core, Wayne R. Guay, and Rodrigo S. Verdi, looking primarily at agencies such as hospitals and charities, finds a correlation between excess endowments and inefficiency of the nonprofit organization.
To be sure, the endowment-expense ratio is a flawed measure; for example, an institution will adjust its operating budget in response to the sizeand changes in sizeof the endowment, and those steps can significantly change the ratio. Nevertheless the wide range of endowment-expense ratios (Raymond Fisman and R. Glenn Hubbard, in 2002, found a range from essentially zero to nearly 100 for large nonprofit organizations) means that this ratio still retains important information about the effective size of an endowment, despite the uncertainties.
With the recent hurricanes in mind, one might argue that an endowment provides an insurance policy, a cushion against catastrophe. But even if you accept that premise, it is still hard to justify an endowment that exceeds the total value of the physical plant of an institution (which is certainly the case for Grinnell, at least). Moreover, many would take exception to the idea that such a cushion is an effective use of an endowment. The very premise of insurance is collective assumption of small risks of catastrophe, obviating the need for myriad individual nest eggs. Even if an institution failed to obtain such insurance in advance, there is no reason to believe that operating in debt while recovering from a catastrophe would be to the long-term disadvantage of the institution. In fact, Henry Hansmann argued in The Chronicle in 2004 that an institution truly interested in intergenerational equity of finances would always operate in debt.
If we accept the endowment-expense ratio as a good measure for comparing institutions, we discover a surprising result. The endowment-expense ratios of Harvard, Princeton, Williams, and Swarthmore are well behind that of Grinnell. On the basis of figures from The Chronicle, Grinnell’s ratio in the 2003 fiscal year was 13.9, compared with 7.5 for Harvard, 10.7 for Princeton, 8.3 for Swarthmore, and 7.4 for Williams.
If we could depend on steady growth and income from an endowment, then determining its ideal size would be elementary: It should be just large enough to pay out at a rate necessary to meet expectations for the budget while expanding the principal at the same rate as the budget. At Mythic College, with an average budget-growth rate of 6 percent and an average endowment-growth rate of 10 percent (after payout), the ideal endowment value in 2005 is $765-million to support the entire college budget. Mythic’s 2005 endowment value is already larger than necessary to support the entire college budget in perpetuity — if we make that crucial but dicey assumption of a constant growth rate.
Of course, those calculations assume that the payout rate is increased considerably: from 4.5 percent to nearly double that. In fact, increasing the payout rate is essential for achieving financial stability or, in other words, keeping a constant endowment-expense ratio. If we assume constant growth rates, we need to consider three essential elements in maintaining the endowment-budget balance: the payout rate, the budget-growth rate, and the endowment-growth rate. The budget-growth rate plus the payout rate must equal the endowment-growth rate (before payout) to maintain a balance between the endowment and the budget. A long-term average growth in endowment of 10 percent after payout — the percentage includes equity appreciation, dividends, and gifts to the endowment — has been typical over the past decade for the institutions mentioned above. Moreover, it is not clear that higher education can get away with many more increases as large as those in, say, the health-care sector of the economy. So the 6-percent budget-growth rate may be unreasonably high in the long term — and if that is true, the ideal endowment value would be even lower.
It does not take a Wall Street wizard to point out that endowment growth is not constant. Most endowments rely heavily on stock equities, so the Dow Jones industrial average is a plausible model of endowment fluctuations. We have recently experienced about five years of a nearly static Dow, and recent history can point to even longer periods of inaction, even though the long-term average growth rate of the Dow over many decades is roughly 10 percent per year. The usual approach to these fluctuations, not surprisingly, is caution, with payout rates set conservatively.
In considering stock fluctuations, Grinnell trustees over the past decade have evolved a system of breaking the college budget into two parts. The base budget contains the normal operating budget of the college, roughly half of it coming from endowment payout. The endowment-payout rate is slightly over 4 percent, which provides $10-million to $20-million more cash than desired to support the base budget. (Roughly 50 percent of the base budget is supported by endowment payout.)
The surplus is used to support other projects, outside the regular operating budget. These funds at Grinnell have come in two incarnations: the now-terminated Fund for Excellence and the Capital Reserve Fund. The former established a number of centers (for prairie studies, for humanities, for international education); the latter has contributed to construction projects (residence halls, athletics facilities, and some academic facilities).
The unstated, or at least understated, goal of these special funds is to add significantly to the institution without incurring significant continuing expenses — by, say, replacing an aging building or obsolete computers. In that way, a sudden drop in the endowment would not put the college in serious financial jeopardy. But the Fund for Excellence solicited proposals from faculty members, and projects that it supported have generally incurred continuing expenses that have been rolled into the base budget. The Capital Reserve Fund has spawned a building frenzy, with two or three major construction projects in progress at any time over roughly the past five years. Some of those projects replaced demolished buildings, but the majority have been net additions to the campus infrastructure, with continuing expenses for maintenance and climate control.
Both of the funds have been qualified successes. Few question that the results have contributed to the institution, although some have argued that the money could have been better spent. The continuing obligations that ensue are not negligible, but not as burdensome as hiring a flock of new tenure-track faculty members. The principal failure, I think, is that each of the funds was only a short-term fix to the problem of turning the substantial endowment income to the benefit of the institution.
Partly with that in mind, the trustees are making long-term plans, which include an expansion of the faculty. But even those plans are neither of the scale (in dollars) nor of the correct character (in terms of flexibility) to deal with the substantial but variable income of the endowment. That failing resulted from a lack of dialogue between those who understand the finances best, principally the trustees, and those who understand the mission of the college best, principally the faculty. During the planning process, the topic of the endowment was implicitly off the table. That is not such a surprise — the board is fearful that the faculty simply wants to spend the endowment, and the faculty doesn’t understand why the board doesn’t want to spend the endowment. But without serious, extended, and inclusive discussions, the adversarial nature of trustee-faculty relations with regard to the endowment will persist.
One could argue for just letting the endowment grow, and spending only what seems wise. Isn’t that the appropriately cautious route? How can too much money possibly be a problem?
Well, it can be troublesome in many ways. It can foster discontent with students in terms of tuition, financial aid, facilities, and staffing. It tests the good will of faculty and staff members toward the institution, and encourages an expectation of pay for anything beyond required service. And it can undermine fund-raising efforts and the corresponding involvement and dedication of alumni to the institution.
Discontent of this sort is most pronounced on the campus. Surrounded by multiple construction projects demonstrating the wealth of the institution, students find it difficult to accept increases in tuition beyond those in the cost of living. At Grinnell that discontent was expressed openly in a student protest at a recent trustee meeting. Department heads preparing budgets are admonished to keep increases comparable to inflation. But they find that advice incongruous when new projects, less directly related to the mission of the institution, seem to be flourishing.
Those problems are possibly outweighed by the stability and budgetary support provided by the endowment. The greater risks, I believe, come from a near-certain drift in mission that mega-endowments engender.
Academic institutions are conservative. Reform and innovation do occur, to be sure, but typically they start from small beginnings in isolated corners. It is rare that a college with a long history will collectively choose a starkly new direction. Faculty members themselves are key elements in this mission molasses; they need time to discuss, argue, and persuade before changes are made.
Once the endowment seriously outdistances its ideal size, however, the ability of faculty groups to design and agree upon changes that will quickly and substantially use the endowment will be insufficient. Those decisions are instead made by an individual or a small group, placing a lot of power in a few hands — those of a president, a cluster of key board members, and a few influential faculty members.
At some point, the money will be spent. I find it inconceivable that those with power over the budget will be able to resist spending excess endowment income, or will switch to lower-yield investments. At some point, too, the IRS could also become involved, enforcing payout requirements similar to those applied to grant-making nonprofit groups. In any event, I can foresee a situation in which apparently cautious policies will lead to an endowment-expense ratio so large that colleges, like cash-rich corporations, will need to make major investments and expansions. Conceivable scenarios include the purchase of the campus of a financially troubled institution in a desirable location, or a massive stake in a distance-learning operation. In any event, the endowment will drive a change in the mission of the institution, a change that will come as a lurch rather than a drift.
If a well-endowed institution is comfortable with its mission, a stable endowment-expense ratio is essential to continuing that mission. Maintaining the balance is easier by far when the budget of the institution does not rely largely on endowment payout. To avoid radical methods of adjusting to endowment fluctuations, an ideal endowment would provide a third to a half of the operating budget, and the endowment-expense ratio would be in the range of five to 10. That range would allow for variation in yield and budget growth rates; for Mythic College, an endowment-expense ratio of 5.9 would provide an average endowment contribution of 50 percent of the operating budget and maintain the endowment-expense ratio. Many well-endowed institutions are already in that range, and I think it is irresponsible of those institutions to postpone inclusive discussions of the pros and cons of moving beyond that level.
Grinnell, with an endowment-expense ratio in 2003 of 13.9, has already exceeded that range by about a factor of two. For Grinnell it is too late to stop at an endowment that plays a more balanced role in the institution. It may even be too late to avoid a sizable change in the character or mission of the college. Major decisions regarding the role of the endowment and potentially new directions for the institution will be made over the next five years.
As a member of the Grinnell College community, I am naturally interested in those decisions, and I hope to do what I can to encourage inclusive discussions that will lead to choices the whole community will endorse. Like many institutions, ours will have to decide how far we should go in balancing institutional identity against layer upon layer of extra financial security.
Mark B. Schneider is an associate professor of physics at Grinnell College.
http://chronicle.com Section: Endowments Volume 52, Issue 39, Page B18
