More On Endowments

Late last week Crooked Timber had a lively discussion about university endowments, prompted by my recent post here and Larry Solum’s response to it. Those who are interested in the topic should take a look at the discussion, as it partially mirrors the debate that is taking place more generally. I’ve been following Crooked Timber with interest, and here’s several points that have struck me:

* I’ll start with the observation I found most interesting: that some elite institutions have a mission that is as much (or even more) about research than about education. I agree that I need to emphasize this distinction more than I have to date. My proposal that an endowment per full-time student of $300,000 or more trigger less favorable tax treatment could penalize institutions whose primary output is research rather than education. Recall, however, that the most frequently proposed trigger is an absolute endowment value of $1 billion or more. Elite research universities tend to have endowments of this magnitude, so my proposal is not tougher on these institutions than the oft-suggested alternative. In fact, my proposed trigger would exempt some research-oriented universities that would otherwise be subject to new tax rules, such as Cornell and Columbia. The institutions most “negatively” affected by the $300,000 trigger are liberal arts colleges with endowments less than $1 billion and small student populations.

More important, however, is that a research-oriented mission actually strengthens calls for increased endowment spending. The sort of research taking place at America’s premier universities is designed to eventually lead to much social good: the easing of the global food crunch, the elimination of certain diseases, and so on, as well as the creation of knowledge more generally. Few science departments, for instance, are likely to argue that a dollar is better spent in the stock market than in their labs. The ability of researchers and scholars to make productive use of endowment funds seems almost endless, as do the potential gains from their work. This strikes me as a strong argument for elite research universities spending more of their endowments than they currently do.

* Any talk of a $300,000 trigger, however, presupposes that at least some universities and colleges should spend more. On this point, some of the comments over at Crooked Timber tipped up what can be characterized as a disagreement about who bears the burden of proof: Congress or universities with mega-endowments. That is, if Congress is considering changing the way it taxes these institutions, do universities bear the burden of convincing us about the wisdom of their endowment spending policies, or does Congress have to convince us that the social good produced by these policies is inadequate given the tax code’s treatment of these institutions and their donors? Because current tax policy so favors universities and colleges (they are exempt from the corporate tax and are not subject to some of the rules that govern other non-profits such as foundations), I tend to think that universities bear the responsibility for demonstrating the wisdom of their spending policies and the relative good created by them.

Regardless of who carries the burden, however, mega-endowments are starting to come under fire for a couple of reasons. One is the cost of tuition, which has skyrocketed. This concern is only partially about tuition at mega-endowment universities and colleges; some of these institutions have been aggressive about controlling tuition and reducing or eliminating it for lower and middle class students. The larger concern, however, is that mega-endowments enable expenditures that are only loosely associated with education or research, such as the construction of student amenities. These sorts of expenditures affect the priorities of less-prosperous institutions, which engage in this sort of spending at the cost of initiatives more directly related to education, including those that reduce tuition. Above all, however, is a belief raised by groups like the Harvard Alumni for Social Action mentioned in my initial post: institutions with mega-endowments could broaden their missions in ways that would require more spending and also produce additional social good.

* As Solum’s post alluded, universities defend large endowments by citing the need for a stable and predictable funding source (“saving for a rainy day”) and by invoking concerns about intergenerational equity. I am sympathetic to the threat of rain, which is why my proposed $300,000 trigger is most likely to affect schools who have 5 years or more worth of reserves, not including the value of their real estate and physical plant. This is also why I would likely oppose Congressional spending requirements aimed at all universities, regardless of endowment size.

Intergenerational equity is the notion that a university needs to spend less so that the endowment can support the same activities in the future as it does in the present. The work challenging the logic of intergenerational equity is not mine; it was done by Henry Hansmann almost 20 years ago. Solum has said that he finds the arguments against intergenerational equity unpersuasive (or at least my recounting of them); I disagree. But don’t take either of our words for it. Instead, check out Hansmann’s work for yourself. The best place to do so is here , although shorter versions have appeared in The Chronicle for Higher Education.

As I studied endowments, I was surprised by how little academic work exists on the subject. I’ve had several people suggest to me that this is because academics fear reprisal from their own institutions. While this may be true in isolated incidences, it doesn’t sound right to me, especially given the politically-charged issues that academics routinely tackle. But I do think that academics need to think seriously about mega-endowments; that’s one of the reasons I urge you to look at Hansmann’s work.

None of what I have written above or in my previous post discusses how institutions might be made or encouraged to increase spending. More on that later.

You may also like...

8 Responses

  1. A.J. Sutter says:

    1. A link to the Hansmann article is here:

    2. I asked the following question in connection with your previous post, but didn’t get an answer, so let me try again in somewhat re-phrased form:

    (a) why should I, as a US taxpayer, feel that this issue is an appropriate priority for Congress at this time? (I am not questioning the “is” of whether Congress is considering this matter, but rather the “ought”.)

    (b) In particular, since you quantify the lost revenue to US Treasury at $18 billion (assuming a corporate tax rate), which is a small fraction of what’s being spent on the war in Iraq, among other misadventures, why ought this to appear on my, or my representatives’, radar now?

    3. Hansmann does make some interesting arguments against intergenerational equity as a justification for huge endowments.

    (a) The most persuasive of these is maybe “[W]hen a university adds a dollar to its endowment for the purpose of an intergenerational transfer, it is implicitly making the judgment that the dollar will have a higher rate of return if invested in stocks and bonds than in educating an undergraduate, or doing research in biophysics, or adding books to the library” (@18). He does make it sound as if they are doing something bad; that merits chewing on to see if there’s some substance under the rhetoric…

    (b) “[T]he university can, in general, only transfer wealth among generations of its own students; it cannot otherwise affect the overall distribution of wealth in society at large” (@18). Considering the research mission of most mega-endowment universities, I’m not sure this one is accurate or persuasive.

    (c) “There is every reason to believe that, in the long run, the economy will continue to grow in the future as it has in the past and that future generations of students will therefore be, on average, more prosperous than students are today, just as today’s students are more prosperous than their predecessors. Thus equity does not call for a transfer of wealth, through saving, from the present generation to the later ones. On the contrary, it would seem more equitable to have future generations subsidize the present” (@14). Here his assumption seems to be unwarrantedly optimistic. In fact, if you give credence to the work of, e.g., economist Nicholas Georgescu-Roegen, his assumption of continued growth is flat-out wrong.

    4. Finally, just a note about the situation in Japan, where I live: the elite universities here are public universities, whose endowments are controlled by the government. There are number of these scattered around the country, including some in poorer, more rural regions. University of Tokyo (a/k/a/ Toukyou Daigaku a/k/a Todai) gets over 50% of all university budget; most of the rest goes to one or two others, such as U of Kyoto. Those “elites” in the poorer regions get much less. Recently, the LDP-controlled government proposed to have so-called “market-based” funding, based on research output, patents, and similar metrics; the result would be that Todai gets an even bigger slice of the pie.

    I’m very, very far from being a libertarian, but even I am skeptical about government regulation of endowments. If you want to encourage universities to give more scholarships, that sounds fine — though it will again attract students to the rich schools, not the lesser ones. Maybe a better solution is to shift spending from some wars here and there, and target it to some of the poorer institutions. I’d like to hear why you disagree.

  2. Sarah Waldeck says:

    As to Sutter’s first comment: Thanks for the link to Hansmann. You will see that I edited my post to include it.

    As to Sutter’s second comment about Congressional priorities and revenue: In my response to your comment to the original post, I mentioned $18 billion as what would have been collected in 2007 if Congress taxed investment income and upwards of $4 billion as what would have been collected if donations to educational institutions were not tax deductible. (These figures, BTW, are from the Government and Finance Division and the Tax Expenditure Budget.) As Professor Evelyn Brody has pointed out, other tax expenditures help create demand for higher education, such as the HOPE tax credit (projected to cost $3 billion in 2008) and the Lifetime Learning tax credit ($2 billion). But you are right that the total federal expediture for higher education pales in comparision to, say, what the government is spending on Iraq.

    But I reject the premise of your question, which implies that because Congress spends more money on other things, it shouldn’t bother with this issue. I approach endowments from this perspective: Congress spends somewhere in the neighborhood of $25 – $30 billion on higher education. As a taxpayer, I’m interested in whether current policy yields the most optimal results, or whether a different policy would yield greater benefits for higher education across the board. (As an aside, I’m reminded a bit of the debate over the estate tax, where some critics dismiss it on the ground that the tax raises so little revenue that we shouldn’t bother with it. I would rather look at the merit of the issue than dismiss it as unimportant because “only” $20-some billion is at stake.)

    The other point is that, notwithstanding the “ought,” Congress “is.” What Congress does or does not do will affect universities, faculty, donors, and students. The “is” provides another reason to care.

    As to Sutter’s third comment about Hansmann, I think the point about research helping to spread the wealth is well-taken. Similar to what I said in my post, however, I think that just strengthens the argument for spending more now.

    4. As to Sutter’s fourth point about Japan, there were a couple of comments over at Crooked Timber about foreign university systems. You may want to take a look at them, if you haven’t already.

    In my estimation, BTW, the best way to change university spending policies is to change donor behavior. I’ll have more on this in later posts.

    As to Sutter’s last paragraph, the world would be a very different place if I could reorganize the Tax Expenditure Budget according to my own priorities! From the tenor of Sutter’s comments, I assume that it would look different under his as well. 🙂

    Finally, Sutter’s comment to the original post had a reference that I understood to be about donor-restricted funds. I’m not sure whether I’ll work that issue into a future post, so I’ll comment now. It is true that some endowment funds are donor-restricted. But this defense of spending policies tends to be overstated. According to Senate testimony by university representatives, about 45% of endowment funds are “restricted,” which is a bit of a slippery term. Institutions expend significant resources cultivating donors and helping to shape their giving preferences. As such, universities have some control over whether a gift is restricted or not. Moreover, cultivated gifts often pay for expenditures the university would have made even without a gift and thereby allow the institution to redirect funds to current expenses or to the endowment.

    Thanks for all your comments, Sutter.

  3. Joey says:

    Is it fair to say you are essentially trying to press universities toward spending on a more immediate time horizion — e.g. spend $1 more today instead of $10 more, 25 years from now? (That assumes a 10% return.) Either way, whatever the school’s payout ratio is, in the end all the endowment funds are going to generate money to fund the work of the university (research, teaching, etc.). Right?

    If I understand your proposal correctly, what is the argument for why society should favor $1 of research/teaching/etc now as compared to $10 of research/teaching/etc in 25 years? What motivates you to think that spending sooner is better in some way?

    If you think there is something better about spending $1 today, then what would be wrong with requiring a very high payout ratio, say 15%, until all the great endowments have been spent on current projects and there is nothing (or, less than $1 billion) for future generations?

    I guess I am just unclear on what motivates your idea that society should have a view about what is the best time horizon to choose.

  4. Sarah Waldeck says:

    To be clear, the intergenerational argument is that endowments have to spend less today so that they will be able to support the SAME amount of educational / research activities in the future. At some point, $10 will buy the same amount that $1 buys today. Universities are not arguing that they need to spend less in the present to be able to support more activities twenty years from now. As other commentators have pointed out, intergenerational equity means that institutions forgo cheaper goods / services in the present for more expensive ones in the future.

    I am not a fan of payout rates as a general matter. In the foundation context, their primary purpose is to ensure that the foundation exists to serve the public good, instead of private interests. I have not heard even the harshest university critics level this sort of allegation at universities. In general, I tend to think that pay out rates are a “one size fit all solution” that will not necessarily serve universities or the public well. I also think that the natural tendency will be to spend at whatever the minimum is, when in some cases it might be appropriate to spend more.

    But your question is about more than pay-out rates. As a general matter, I support mega-endowment institutions spending down their endowments to the point where their endowment-expense ratio is about 5:1. (As a practical matter, however, I am skeptical about whether some of the wealthiest institutions can broaden their missions enough to accommodate this level of spending.) On the flip side, I support less-wealthy institutions building up their endowments to about this level. I think endowments provide important stability; this is why I’ve said that I am sympathetic to rainy day arguments.

    Time lines are significant because institutions and their donors receive the favorable tax break in the present. On the most fundamental level, Congress provides favorable tax treatment to non-profits because they provide benefits to the public that government might otherwise have to provide. There is some immediacy or least near-term calculus here. Neither non-profits nor universities exist to serve as financial repositories. Of course, there is a point at which a particular level of spending can de-stabilize an institution, or at which a particular institution ceases to get bang for its buck. But no mega-endowment university has convincingly argued that it has reached (or is close to reaching) this point.

  5. F says:


    Your paper was full of interesting and useful information, and I tend to agree that limitations on endowments may have some utility. Unfortunately, I think you have chosen the wrong measure. You yourself state that the endowment/expenses ratio is more relevant, but you discard this measure for trivial reasons in favor of the endowment/student measure. The biggest differences between rankings using these measures is, in fact, the research institutions, whose expenses per student are much larger. Therefore, by choosing the endowment/student ratio, you are unfairly (and quite lopsidedly) penalizing the research institutions.

  6. Joey says:

    To be clear, the intergenerational argument is that endowments have to spend less today so that they will be able to support the SAME amount of educational / research activities in the future. At some point, $10 will buy the same amount that $1 buys today. Universities are not arguing that they need to spend less in the present to be able to support more activities twenty years from now.

    I find this turn in your argument interesting, but ultimately a little unsatisfying for the following reason. Large university endowments earn a return much greater than the rate of inflation — indeed, even greater than the extra-high rate of inflation in the cost of research, teaching, etc., measured in most of the ways it can plausibly be measured. Therefore, the choice is not between one real dollar of research today and one real dollar of research in the future. By saving and investing, substantially more research and teaching can be had in the future, even after adjusting for inflation. The intergenerational choice is between less now and more later.

    You characterize universities as “not arguing that they need to spend less in the present to be able to support more activities twenty years from now.” I have no reason to doubt your statement that universities are not making this argument. But it seems to me that one of the strongest arguments for growing an endowment is that the most well-endowed universities are consistently stunning investors, who can do much more in absolute terms by spending later rather than now.

    To put it even more starkly, as university endowments grow over time, total revenue streams dedicated to research and teaching grow in absolute terms, increasing the real amount of money going to research and teaching in the world. You raise various other important issues in your paper and subsequent comments that my claim here does not address (such as the effect of inequality in endowments on the majority of institutions with smaller endowments) but I think you should at least have to answer the objection that cutting endowments to produce more research and teaching now will reduce the total spending on research and teaching for all future time periods — and those losses will be greater in absolute terms than the short-term gains.

  7. Sarah Waldeck says:

    In reponse to F:

    As you recognize, I agree that the endowment-to-expense ratio is the most revealing measure of an endowment’s true size. Earlier versions of the paper endorsed this measure, largely because of its accuracy. In the end, however, I became convinced that use of the ratio was too complex for regulatory purposes.

    I think it’s important to keep the measure simple. I can easily imagine proposals to make, say, spending requirements apply to all institutions, regardless of endowment size. (As I have stated previously, I would not favor this.) But those arguing in favor of universal applicability would likely cite such complexity as an argument in their favor.

    In response to Joey:

    Your point is well-taken. Whether it is ultimately persuasive depends on the probable gains from having more research or education occur in the present. This is obviously much more difficult to quantify than gains from financial investments. Over the long-term, would we be better off spending more on, say, genetic research now and sacrificing whatever above-inflation yield we would have received from the dollar that was invested in such research (even assuming those investment gains would be invested in research 50 years from now)? I will give this point more thought. Thank you.

  8. B says:

    Great post, v interesting and informative.I found a debate on university endowments in Britain here: