Forcing foundations to give more cash grants each year may cheat future generations of their just desserts. 

Bill Gates has said that when an AIDS vaccine is produced, the Bill and Melinda Gates Foundation will fund the vaccine’s distribution around the world even if the foundation has to spend down its $24 billion endowment. For now, and until the vaccine is found, however, the foundation is distributing funds at about the legally required rate of 5 percent per year. 

Recently, strong arguments have been made to foundation managers and legislators that foundations should distribute their assets at a faster rate. In September, Congress seriously considered a measure that would have increased annual foundation payouts before putting it aside. The foundation payout debate has spanned decades, but advocates for a faster payout received a boost last year from an article by McKinsey & Company consultants Paul J. Jansen and David M. Katz, “For Nonprofits, Time is Money,” in the McKinsey Quarterly. They argued that we should view foundation grants as an investor would view an investment. Just as investors would choose to receive a dollar today rather than a dollar a year from now, so too is a dollar of charity given today worth more to society than a dollar of charity given in the future. That position picked up more publicity when former New Jersey Senator Bill Bradley, now a consultant to McKinsey, joined Jansen in making the same argument in a New York Times op-ed, “Faster Charity,” on May 15, 2002. 

If Bradley and the McKinsey authors are right, the Gates Foundation should reassess its strategy. The foundation would need to discount the social benefit of a future AIDS vaccine to a “present value,” just as an investor would discount future investment returns to present value. This discounting exercise would reduce the vaccine’s value to a fraction—very likely a small fraction—of the benefit that the vaccine will produce when it is actually distributed. Thus, more immediate grants to charity would appear more socially valuable in comparison, and to that extent, the McKinsey authors argue that foundations should accelerate their payout rates. 

But the McKinseyites are wrong. Their discounted cash flow approach doesn’t fit the foundation payout issue. Under certain circumstances, there are good reasons for foundations to favor higher payout rates, and perhaps even good reasons for the law to mandate them. The time value of money, however, isn’t one of them.

Current Versus Future Charity

The issue of foundation payout rates comes down to a tradeoff between charity for the current generation and charity for future generations. The lower the payout rate, the greater the amount saved and invested, and hence the greater the amount that can be distributed to future generations. Conversely, the higher the payout rate, the lower the amount available for future distribution. The arguments of those who advocate higher payout rates amount to arguments that foundations should provide more money to current charity and less to future charity. Foundations that resist higher payout rates are in effect arguing for more future charity at the expense of current charity. 

The McKinsey authors argue that a foundation dollar distributed to charity today is worth considerably more than a foundation dollar distributed in the future. Consequently, they say, foundations can increase the value of their grant making by increasing their payout rates. They explain that they apply “a standard financial concept known as the ‘time value of money’” to reach this conclusion. This is the same “discounted cash flow” approach that corporations and investors use in deciding whether a current investment is justified by its projected returns. To evaluate a proposed investment, a company projects the investment’s future cash flow and discounts it to present value using a discount rate that reflects the company’s cost of funds or the rate of return the company can earn from an alternative investment. An individual investor would approach an investment the same way, but in addition, an individual would consider whether he prefers immediate consumption to the opportunity to consume more in the future when the investment pays off. 

The McKinsey authors are not the first to apply the discounted cash flow approach to foundation payouts. The U.S. Treasury Department and Congress implicitly took this approach in the 1960s when the payout rate was initially enacted. They were troubled by the fact that a donor to a foundation takes a tax deduction at the time of the donation but the donated funds might not reach actual operating charities until many years later. Congress and Treasury believed that because of this delay, donors were getting a tax benefit worth more than the charitable benefit they produced. Other advocates for higher payout rates have referred to the time value of money as well. 

The McKinsey authors, however, provide the most detailed explication of this argument. They begin their analysis by constructing a hypothetical foundation that will exist for 50 years. [See chart.] Their foundation begins with assets of $1,000, it earns a 10 percent annual rate of return on its investment portfolio, it incurs administrative costs at the rate of 1 percent per year, and it distributes 5 percent of its assets per year in grants to charity. With these numbers, the foundation makes grants of $50 in its first year. In its 50th year, its assets will be more than $5,000 and it will make grants of $257. The foundation’s grants over 50 years will total $6,355. 

Sounds like a valuable social institution to me, but the authors are not so sanguine. They calculate the present value of the foundation’s grants to society by discounting the 50-year stream of grants at two alternative rates: the 10 percent rate that the foundation earns on its investment portfolio, and a 15 percent rate that they say the foundation could earn for society by making grants today. Running these calculations, the authors find that the foundation’s $6,355 in grants over 50 years is actually worth less to society than the $1,000 with which the foundation started. (It is worth $830 using a 10 percent discount rate and $500 using a 15 percent rate.) They run various scenarios through their spreadsheet to show that foundations that want to increase their value to society should increase their payout rates above 5 percent. But they neglect to point out that under their valuation approach, the best a foundation can do is break even in terms of creating social welfare—and that, with the 15 percent discount rate, the only way a foundation can do even that well is to distribute 100 percent of its assets immediately and to do so without incurring any administrative costs. 

The McKinsey authors explain that skilled grant making can offset the ravages of time on a foundation’s social worth, but holding the quality of grants constant, their point is simple: Future charity is worth less than present charity, and it is the time value of money that makes the difference. 

The McKinsey authors’ analysis is simply arithmetic. By assuming a social rate of return on a foundation’s grants (15 percent) that is higher than the rate of return on its investments (10 percent), their calculations would lead the foundation to distribute all its funds immediately. But if a foundation’s grants yield only a 9.9 percent return for society, then those same calculations would lead the foundation to invest its cash forever and never make a grant! Something must be wrong with this approach.

The Inapplicability of the Discounted Cash Flow Approach

The McKinsey authors’ analysis is flawed, not merely because of the numbers they use, but because the discounted cash flow approach is not an appropriate method for measuring the value of foundation grants made in the future. When the McKinsey authors measure the present value of their foundation, the value of the grants that the foundation makes each year is divided by a discount factor. 

For example, take the value of grants made at the beginning of the 48th year—$241. Divide that by 1.15 to the 47th power, or 712. (Using the 10 percent discount rate the figure would be 1.10 to the 47th power, or 88.) The result is that the $241 in grants for year 48 is worth about 34 cents. 

Accordingly, if a grant will be made 48 years from now to fund a soup kitchen serving three meals a day for the full year (a total of 1,095 meals), the present value of that grant is just one and a half meals—brunch! If the foundation had a choice of serving brunch today or three meals a day for a full year in 48 years, the discounted cash flow approach would tell us it is a coin toss. This low valuation of the soup kitchen’s services is based solely on its clients’ hunger occurring in the somewhat distant future rather than today. 

There are several reasons why the discounted cash flow approach is irrelevant to the foundation payout issue. 

Most fundamentally, by discounting future grants to present value, we would be saying that future grants are worth less to society than current grants. Using the soup kitchen example above, a grant of $241 in 48 years is worth a lot less than a grant of $241 today; using a 15 percent discount rate, it is worth only 34 cents today. But why? In the private investment context, if investors can earn 15 percent on their money, they can convert the 34 cents into $241 in 48 years. To an investor, therefore, receiving 34 cents today and receiving $241 in 48 years are equivalent. 

But when we compare a grant to charity today with one made in 48 years, we are comparing the benefit of helping one group of people today with the benefit of helping another group in 48 years. There is no similar equivalence. Why would 34 cents worth of food to a group of hungry people be worth the same as $241 of food to a different group of hungry people simply because the two groups live at different times? By invoking the discounted cash flow approach, the McKinsey authors have adopted what economists refer to as a “pure time preference” in allocating resources over generations. Such a preference is difficult to justify as an ethical or economic matter. Frank Ramsey, who in 1928 was one of the first economists to analyze resource allocation over time, called the discounting of funds allocated to future generations “ethically indefensible and aris[ing] merely from the weakness of the imagination.” 

Secondly, there is no basis for discounting a future grant at the rate of return a foundation earns on its investment portfolio—the 10 percent rate the McKinsey authors use. In the private investment context, the projected cash flows of a proposed investment are discounted at the rate of return available on an alternative investment; by making the proposed investment the company or the individual would forgo the alternative investment. [See sidebar, p. 17.] A foundation, however, doesn’t lose an opportunity to earn a return on its investment when a grant is deferred. On the contrary, as the McKinsey authors recognize, the funds remain invested in the foundation’s portfolio, earning a 10 percent return. This step in the authors’ discounting exercise amounts to inflating and deflating the foundation’s assets at the same rate, which results in a wash—there is no loss of value as a result of delay regardless of the payout rate. 

The reason the McKinsey authors find that the value of the foundation’s grants is less than the $1,000 with which the foundation started is because their foundation incurs administrative costs in making grants. Although they recognize that skilled grant making can produce social gains, their calculations include only the cost of grant making, not the benefit. In the McKinsey authors’ calculations, even a penny of administrative costs would render the foundation a net loss to society. The presence of administrative costs, however, is not a per se reason to increase payout rates. 

Third, the McKinsey authors are correct in recognizing a social opportunity cost in forgoing earlier grant making. The cost of that lost opportunity is the “return” that society could have reaped if the foundation had made grants earlier. The authors recognize that social returns are hard to quantify and that selecting a discount rate is difficult as well, but they select a 15 percent social rate of return as “a conservative estimate for the upper end of our range of rates.” They base this claim on work done by the Roberts Enterprise Development Fund (REDF) to measure the impact seven nonprofit organizations had in running business enterprises that train and employ an inner-city population. This figure, however, is not representative of the social returns generated by the entire charitable sector, which includes art museums, prep schools, soup kitchens, hospices, universities, and innumerable other sorts of organizations. Moreover, although foundations are commonly interested in making grants that will produce a return to society that continues for some period of time, many grants—to the opera, symphony, soup kitchen, and homeless shelter, for example—produce benefits that are better characterized as largely consumption rather than investment. 

Fourth, even assuming that a grant yields a social return—of 15 percent or whatever—the McKinsey authors’ application of the discounted cash flow approach assumes that this return will be maintained over the long run—50 years in their hypothetical foundation. When one applies such a discount factor to a grant to be made 50 years from now, one says that the money could be invested today to generate a 15 percent return for 50 years. At that level of sustained social gain a grant of $100,000—say, to fund a college scholarship for at-risk youth, or to support the local symphony—would yield $108 million worth of gains to society at the end of 50 years. This seems unlikely, and it certainly has no basis in REDF’s experience. 

Fifth, even if a current grant to charity does yield a longterm social return, unless the return continues in perpetuity, applying a discount rate to future charity gets us back to the problem with which I began this analysis: the favoring of one set of beneficiaries over another based simply on the period of time during which they live. As I discuss below, there may be justifications to such a preference, but they are not found in the discounted cash flow analysis. 

Finally, if the discounted cash flow approach were applicable to the timing of grants, it would be applicable to the evaluation of grants themselves. To evaluate a grant, a foundation manager would discount its projected social return. The discount rate, at a minimum, would have to equal the rate of return earned on the foundation’s investment portfolio—10 percent in the McKinsey authors’ hypothetical. This would lead a foundation to forgo grants that are expected to yield social benefits, if those benefits are less than the expected financial return on the foundation’s investments. In other words, if grants to a soup kitchen or an opera or a school are not expected to yield what the bank or the stock market will pay, the foundation should not make the grants. This surely is not a proper comparison. To compare the private return available in the market with the social return available in the charitable sector is an error of the apples-and-oranges variety.

Similarly, if a foundation were to follow the discounted cash flow approach, it would have to discount the projected social returns from one grant by the social returns available on other grants. Foundations already do this implicitly when they compare two grants in the same field. But the discounted cash flow approach takes it a step further. If, for instance, a foundation funds research on the history of Western civilization, the discounted cash flow approach would require the foundation to discount the projected social returns from that research by the social return it could achieve with a grant anywhere else in the charitable sector. Such a practice would sacrifice diversity in grant making. 

Balancing Current and Future Charity: A Fresh Start

So if the discounted cash flow approach is not useful, how should foundation managers think about the tradeoff between current and future charity? 

The tradeoff between current and future charity is a version of a problem with which policymakers, economists, and philosophers grapple when considering very long-term public investments in energy production and environmental protection. How much sacrifice should the current generation make so that future generations can have a cleaner environment, cheaper energy, better health, and longer lives? The question for foundations is similar. How much charity should we withhold from the current generation to provide more charity for future generations? 

The challenge of how to allocate resources among generations is fundamentally an ethical question, with economics helping to highlight the tradeoffs. One realization that has come out of the debate over long-term public investment is that the pure timing of a social benefit— whether this generation or a future generation enjoys the benefits—should be irrelevant to its social value from either an ethical or economic perspective. So, for instance, if greenhouse gas regulation today improves the lives of people living 100 years from now, the mere fact that the benefit will be enjoyed by people living so far in the future doesn’t make its social value smaller. The same is true of the future benefit that comes from a foundation’s decision to adopt a low payout rate today to support charity in future generations. There may be a temptation to care more about the current generation than about faceless generations in the future. But the philosopher John Rawls observes, “The different temporal positions of persons and generations does not in itself justify treating them differently.” 

Those advocating higher payout rates legitimately point to a dire need for current charity. As a matter of advocacy, this approach is understandable. But as a matter of analysis, we need to recognize that current charity comes at the expense of future charity, and that the mere timing of a generation’s presence on this planet is not relevant to the social value of charity provided to that generation. Moreover, because charity deferred to the future earns a return in the foundation’s investment portfolio, a dollar withheld from the current generation can be expected to yield more dollars of charity for future generations. Ben Franklin appreciated this aspect of the tradeoff and chose to hold off giving a few thousand dollars to Boston and Philadelphia in 1790 so that his gift would amount to several million dollars in 1990. [See sidebar, p. 14.] This is surely not to say that we should sacrifice all current charity for the future—in perpetuity. The challenge is to find an approach for analyzing the tradeoff between current and future charity. 

That tradeoff presents three issues for a foundation to confront in determining how much to save and how much to give. The first two reflect the goal of maximizing aggregate social welfare across generations. The third reflects a goal of intergenerational equity—a notion that there is a limit to what we can ask one generation to give up in favor of another generation for the sake of maximizing total welfare. 

The first issue that a foundation should consider in setting a payout rate is how cost-effective a grant to current charity would be, compared with future charity, in providing a charitable service. Despite the fact that a dollar of today’s charity comes at the price of many dollars of future charity, certain kinds of charity today will be more cost-effective; current and future generations will be better off if these charitable services are provided sooner rather than later. For example, if a foundation’s goal is to preserve open space, doing so sooner may be better than doing so later, when the choice of open space to preserve will be more limited. The same may be true of efforts to reduce population growth in an overpopulated region or to cure an infectious disease. It may be true as well of some educational programs, but only if one expects the benefits of current education to have indirect effects on the descendants of current students in perpetuity. If current charity in areas such as these, and surely others, produces benefits that compound in perpetuity at a higher rate than assets in the foundation’s portfolio, then not only will the current generation benefit from a grant today but future generations will be better off as well. 

Before salting its money away for future generations, a foundation should also ask itself whether future beneficiaries of its mission are likely to be better off than current beneficiaries. For example, perhaps with continued economic growth over the generations, art aficionados of future generations will be wealthier than the art aficionados of today. If so, there is less reason to save today to support the arts in the future. In addition, economic growth over the generations is likely to mean more donations to the nonprofit sector in the future. More immediately, some expect a massive flow of funds to the nonprofit sector as the baby boomers pass on their wealth over the next 20 years. If the charity sector of the next generation will have more funds than the sector has today, then there is less need to sacrifice current for future charity. Or perhaps the needs that a foundation serves will be less severe in the future. A problem may be solved, or a service now in short supply may be abundant. If, for any of these reasons, future generations of charitable beneficiaries are expected to be better off than the current generation, then a foundation should put a thumb on the scales of the current generation. This does not amount to discounting the future generation because it will arrive on the scene in the future. Rather, it is a matter of giving resources to those who are worse off rather than those who will be better off. 

The third issue, intergenerational equity, provides a basis for a foundation choosing to give a dollar of charity today rather than more dollars in the future. In contrast to the first two, this is not a matter of maximizing welfare across generations. It is a potential reason to favor the current generation at the expense of future generations. This principle weighs against the goal of maximizing aggregate welfare in the charity sector over the generations. As an ethical matter, there must be a limit to the extent of sacrifice any generation can be asked to make for future generations, even if further sacrifice would lead to net gains in the future. 

There also may be situations in which certain members of the current generation have a particularly strong ethical basis for deserving something more than members of future generations (and more than others in the current generation). Innocent victims of a war waged by the current generation, for example, may have an ethical claim to funds that the current generation has accumulated for charity. Ideally, each foundation would strike a balance between equity and wealth maximization, as it deems appropriate for society as a whole. Just as foundations distribute their funds across the charitable sector as they choose—focusing on maximizing social returns or on other ethical considerations— they should do the same with their distributions over time. 

So how does the Gates Foundation’s AIDS strategy look under this approach? First, Gates should not worry about discounting the value of lives saved in the future as a result of an AIDS vaccine. The foundation’s strategy should be analyzed based on the cost-effectiveness of providing less now and more later to combat the disease. Delivery of the vaccine, even to the next generation (or the one after that), may well be more beneficial to society over time than adding yet more Gates Foundation funds to the AIDS effort today. This is the type of judgment that individual foundation donors and executives must make. If there is a flaw in the Gates Foundation strategy, it is that there may be more philanthropic dollars available to support the delivery of an AIDS vaccine when it is developed. At the margin, Gates Foundation funds may better the lives of more people if used sooner. That, however, is also the type of judgment that must be left to foundation donors and executives.

The Mandatory Payout Requirement 

If there is not necessarily a downside to society for a foundation to favor future charity over current charity, then why do we need a mandatory payout rate? It is not because there is a mismatch, as Congress believed in 1969, between the value of the tax deduction and the value of charity given in the future. As explained above, future charity is not necessarily worth less than current charity. Furthermore, since future charity benefits from compound growth, the tax deduction for a donation to a foundation is equal to the present value of the foundation’s future grants. 

So why not allow foundations to distribute less now and defer more resources to future generations? 

If foundation managers were guided entirely by social welfare considerations in setting their payout policies, then no minimum payout law would be needed. Foundation managers, however, seem to be influenced by the prestige associated with large endowments, and foundation donors seem to be influenced by notions of immortality associated with perpetual existence. Consequently, donors and managers seem to have a personal bias toward lower payout rates. 

The minimum payout requirement responds to this self interest in a fairly moderate way. It basically allows a foundation to maintain its principal and to make grants in perpetuity at the 5 percent rate. This allows donors immortality and forces foundations to treat current generations at least equally with future generations. In the long run, the minimum payout requirement is expected to hold foundation endowments constant, so it inhibits foundation executives from vying for prestige by growing their endowments. 

Should the minimum payout requirement be increased as some advocates have urged? Thisdependson the behavior of foundation managers. If the personal biases of foundation executives play too strong a role in allocating funds between current and future charity, an increased payout requirement might be reasonable. Of the factors discussed above, the one that might push in favor of a higher payout requirement across the board is the comparison between the resources and needs of charity today versus charity in the future. If, assume expect, there will be a large flow of funds to the charitable sector over the next 20 years, and if we expect society to grow wealthier and charitable donations to increase with wealth over the generations, then perhaps foundations should anticipate this new money and devote more funds to current charity. On the one hand, from a policy point of view, there is no problem with foundations spending themselves out of existence as new foundations take their place. On the other, increasing the payout requirement to a level that would place foundations’ perpetual existence in substantial jeopardy could make the establishment of foundations less attractive to donors, which could result in less charity for present and future generations. 

Reaching a conclusion about whether to up the payout rate will involve considering a host of factors, but not the one that’s been getting the most attention: discounted cash flow analysis. That approach amounts to a pure preference for the current generation over future generations with no ethical or economic basis. 

The allocation of charity across generations must be understood in much the same way as the allocation of environmental resources across generations. There may be justification for favoring current charity over future charity in some situations, but not as an absolute matter.