3
Modifying Endowment Spending Rules: Is it the Cure for Overspending?
Roger T. Kaufman
Smith College
Geoffrey Woglom
Amherst College
March, 2005
Abstract
In this paper, we analyze the dynamics of endowment spending and real endowment values using rules that tie endowment spending to inflation. Numerical examples demonstrate that under a pure inflation rule, spending rates will tend to drift away over time from the appropriate rate, leading to either rising or falling real endowment values. Under a banded rule, the drift is limited by the upper and lower limits of the band, but spending rates tend to get stuck at these limits, again leading to volatility in real endowments. The Yale/Stanford, mixed rule, a rule based both on asset values and inflation, avoids these difficulties. Monte Carlo simulations confirm our analysis in the case of uncertain and volatile asset returns.
Keywords: educational finance; educational economics.
3
Introduction
The costs of higher education are an on-going concern to both parents and policy makers. Many are perplexed that elite institutions simultaneously charge high and seemingly ever rising tuition rates at the same time the institutions are amassing substantial endowments. What is the purpose of these endowments and how do colleges and universities decide on how much to spend out of endowment? These questions are particularly relevant now because of the recent turbulence in financial markets. The bursting of the tech bubble has led many colleges and universities to reexamine their spending policies in the light of falling endowments. This paper is concerned both with the principles determining endowment management and with recent proposals for altering spending policies. We begin with an overview of college endowments and endowment policies before turning to the specific proposals for alternative, endowment spending policies.
Endowment Facts, Principles of Management, and Recent Performance
Endowment Facts:
As Henry Hansmann (1990) noted, American colleges and universities have a unique organizational form. Unlike organizations that are organized to realize a profit and which finance their activities with borrowed capital, either debt or equity, private colleges and universities maintain large reserves of financial wealth, or endowments. To get a sense of what we mean by “large” it is helpful to turn to the statistics on endowment as gathered annually by the National Association of College and University Business Officers, or NACUBO. In the 2003 NACUBO Endowment Study, 717 public and private participating institutions reported total endowment wealth of over $230 billion.[1] At these institutions total endowment wealth is both substantial and highly skewed. The 39 most affluent institutions control 57.8 percent of total reported endowment wealth, or over $133 billion. Of course, the wealthiest institutions tend to be large, but the disparities in size do not explain the disparity in endowment wealth. Among the 29 most affluent private institutions, endowment wealth per full time student amounted to almost $331,000, as compared to the average of over $35,000 per student among all participants. College and university endowments are often large and play a substantial role in higher education finance, particularly at the wealthiest private schools.
There is evidence that the disparity in endowment wealth is growing. Kaufman and Woglom (2005) studied the finances of elite liberal arts colleges between 1996 and 2001. They found that the wealthiest schools in their sample spent more on their students while charging similar comprehensive fees. Yet the wealthiest institutions still managed to add to endowment wealth at significantly greater rates because of their disproportionate initial wealth. While large universities have not explicitly been studied, it is strongly suspected that this same pattern of growing wealth disparities would characterize elite private universities. Endowments are large, particularly at elite schools, and the disparities among institutions are growing.
Purposes and Principles of Endowment Management
What one makes of these facts depends on one’s view of the role of endowments, and surprisingly the precise purpose of endowments has never been articulated, either by academics or practitioners. Hansmann (1990) examined a range of possible reasons for endowments (e.g., endowments as: financial buffers, a means of preserving traditions, a way to insulate the university from outside demands), but finds all of these reasons to be unpersuasive. He argues that endowments may reflect the particular perspectives of trustees. Trustees may have difficulty measuring the output and achievement of their institutions, and therefore may be tempted (particularly given the business background of many trustees) to measure the success of an institution in terms of the dollar value of the endowment. Therefore a college trustee may perceive the “job” to be first and foremost preserving, or better still increasing the value of the endowment.[2]
While Hansmann describes the confusion over the justification for endowments in the first place, there is more agreement (at least, apparently) over the principles for managing an existing endowment. This agreement stems from James Tobin’s (1974) concept of intergenerational equity:
The trustees of an endowed institution are the guardians of the future against the claims of the present. The task is to preserve equity among generations. The trustees of an endowed university like my own assume the institution to be immortal. They want to know, therefore, the rate of consumption from endowment which can be sustained indefinitely. (p.427)
While Tobin does not address Hansmann’s concerns about the initial purpose of the endowment, he does provide a principled argument for why trustees should be concerned with preserving the value of the endowment. Unfortunately, while the concept of intergenerational equity is appealing,[3] its practical implications are less obvious. What does it mean to preserve equity across generations when a university’s costs rise faster than the rate of inflation or when the educational mission of the institution expands to cover new areas of knowledge and new academic disciplines? Furthermore, how should the expectation and arrival of new gifts affect intergenerational equity? Consequently, while Tobin’s principle is widely cited, its practical implications are not obvious. Nevertheless, many colleges and universities may appeal to Tobin’s principle to justify an operating procedure in which endowments are managed to minimize a loss in purchasing power.
The Practice of Endowment Management:
Managing the endowment comprises two parts: 1) investment management, viz. deciding on how the investment pool should be allocated into different asset categories; 2) spending management, viz., deciding on how much of the endowment can be spent. This paper is primarily concerned with spending policies at private institutions.[4] At first glance, the problem of spending management may seem superfluous. After all, some endowment funds (called “true” endowment funds by NACUBO) have restrictions on how much can be spent and on what purposes, and these funds comprised over 60 percent of total endowments. But this still leaves a substantial portion of unrestricted funds, implying that colleges and universities have substantial discretion over spending from the unrestricted endowment. For example, in 2003 so-called “quasi-endowment” comprised over 30 percent of total endowment funds among the NACUBO survey participants. Quasi-endowment represents gifts and accumulated income that the institution chooses to treat as permanent capital. Therefore, both the income and principal of quasi-endowment can be spent at the institution’s discretion. Given the size of quasi-endowment and other unrestricted funds, colleges and universities can be viewed as in fact deciding on how much and to what purposes to spend unrestricted endowment funds.
At most institutions the trustees, in consultation with the administration, establish rules about spending from endowment and endowment management. Although these rules can be modified over time, they are useful because they reduce the cost of renegotiating spending procedures every year and provide a source of consistency through time. The rules also serve as an automatic restraint to withstand current faculty and student pressure to devote an inordinate share of endowment wealth to current spending. Finally, rules also provide some protection against “over-reacting” during unusually good or bad economic periods.
Until the late 1960s the endowment-spending rule was relatively straightforward at most educational institutions. Typically, all of the interest income from bonds and dividend income from stocks were spent and all capital gains were allowed to accumulate. Indeed, until the passage by most states of the Uniform Management of Institutional Funds Act in the mid 1960s, many endowment managers were legally prohibited from reducing “principal” for income needs. Following the publication of an influential report by the Ford Foundation (1969) many colleges and universities adopted their current systems in which a constant fraction of the market value of the endowment, called the takeout, payout, or spending rate, is typically allocated to the operating budget each year[5]. This typically leads to spending more than interest and dividends and is set with the intention to spend some of the long-term capital gains on the endowment portfolio. In 2003, the NACUBO Endowment Study reported that over 95 percent of reporting institutions had some form of spending rule, with over 82 percent using a “moving-average rule,” where spending from endowment is based on a pre-specified percentage of a moving average of past endowment values. The popularity of “moving-average rules” has been growing over recent years. In 1999, 72.9 percent of survey respondents used this rule, and in 1995 only 59.2 percent used the “moving average rule.”
Recent Endowment Performance and Controversies over Spending Policies:
Recently, there has been an upsurge of interest in appropriate spending policies and spending rules because of the turbulence in financial markets. The reasons for this interest are apparent in Figure 1, which displays endowment performance for the most recent available data along with the average performance over the preceding 5 years, 1993-98. The two measures of endowment performance shown are the average rate of return on NACUBO endowments and the average growth rate of NACUBO endowments.[6] It is important to distinguish between these two rates. The growth rate of the endowment depends not only on the rate of return, but also on how much of the endowment is spent and on the rate of new gifts to the endowment. Finally, Figure 1 also displays the rate of return on the S&P 500 (including dividends) as a reference.
The decade of the 90s was a prosperous one for endowed institutions. High asset returns allowed endowments to grow at nearly double-digit rates, much higher than general inflation rates. But in 2001, this rosy scenario changed dramatically with the burst of the tech bubble on Wall Street. Because most colleges and universities hold diversified portfolios of stock, bonds and alternative assets[7], the decline in the average rate of return on the endowment was less than that of the Standard and Poors 500, but this still led to substantial losses in average endowment value. In contrast, average spending rates during this period do not show much variation. During the 1993-98 period, spending rates averaged 5 percent of the endowment; they then fall to 4.7 percent during 1999 and rise fairly continuously to 5.4 percent during 2003. Figure 1 provides a summary of the overall problem with spending that has concerned, if not alarmed, many trustees and administrators. While the changes in the average spending rates were not great, they tended to exacerbate the effects of negative rates of return.
Obviously, individual experiences differ, in some cases dramatically. To get a sense of these differences we looked at the percentage growth in endowment values between 2000 and 2001 and then again between 2001 and 2002 among the 25 wealthiest institutions, defined as those with the greatest endowments. In the earlier period, 20 of the wealthiest 25 schools suffered a fall in endowment values, with one school losing 14.2 percent of its endowment during that fiscal year. The following year, 22 of the 25 schools suffered losses in endowment, with 3 schools recording double-digit losses.
Given trustee concern for maintaining the purchasing power of the endowment, recent losses in endowment principal have led many to reexamine spending policies. Some commentators have argued that many of the recent difficulties could have been avoided. For example, Sedlacek and Clark (2003) attribute some of the recent problems to the “overspending” of the late 1990s by colleges and universities, and they believe that this overspending was caused in part by “moving-average” spending rules. They argue that endowment spending should not be tied to the value of the endowment, but instead should increase with the rate of inflation. In this way, much of the increased spending during the 90s and some of the loss in endowment value would have been avoided.
With the hindsight of the 2001-2003 experience, some restraints on spending growth would have been appropriate. But many of the proposed spending rules appear to be implicitly based on an extreme form of intergenerational equity that attempts to minimize the probability of a loss in the real value of the endowment.[8] We believe that minimizing the probability of a loss is a misreading of the principle of intergenerational equity. In a world of uncertain returns, trustees must recognize that the value of the endowment will occasionally fall. If one is unwilling to let the endowment fall in years of low endowment returns, spending will on average be too little and the generations that experience low levels of spending during the bad years will suffer. Put another way, one must recognize that it is possible to spend too little as well as spend too much, and there are losses associated with both mistakes. In the case of spending too much, the loss is obvious: the real value of the endowment falls and future generations will receive less real support. But if one is serious about intergenerational equity, one must recognize the loss of spending too little: the current generation of students will receive less real support than future generations. Trustees may congratulate themselves in the latter case because the endowment is rising in value, but this is in fact a mistake, at least with regard to equity for all generations.[9]
It is time to consider alternatives to “moving-average” spending rules. In the next section, we review how changes in endowment depend upon the return on the endowment and on the spending rate from the endowment. Three endowment-spending rules are then described in detail: two that are tied to inflation and the Yale-Stanford mixed rule, that ties spending to both inflation and the asset value of the endowment. We show how these rules affect the effective spending rate over time and evaluate them in light of the goal of intergenerational equity and the potential problems of spending “too little” as well as “too much.” A pure inflation rule is likely to lead to a spending rate that diverges over time from its initial value. This tendency for the spending rate to drift leads to increased volatility in the value of the endowment. In section 3, we provide numerical simulations of the spending rules. We then illustrate the increased volatility of the endowment under the inflation rule and compare the performance of the inflation rule to that of the mixed rule. In the final section we present our conclusions.