An Economic and Ethical Approach to Charity and to Charity Endowments

Abstract

We examine how and why donors divide gifts between people in the present (across distance) and between the present and future (across time). US donors tend to give less to charities that benefit the poor and more to charities that benefit the non-poor (such as museums, universities, and arts organizations). Many of these wealthier charities have created endowments that benefit not only present persons, but also future persons. We develop a shorthand framework for linking time to distance in charitable allocations that incorporates a “proximity preference,” i.e., charity that prefers those who are nearer to us whether by reason of physical distance, psychic-identity, or temporal distance. Even though ethical considerations suggest that recipients' level of need should be the dominant factor in allocating gifts, donors also express preferences, ceteris paribus, for benefits arriving sooner rather than later, and for recipients who are ``closer'' rather than farther away.

Keywords:

• charity
• endowments
• time preference
• proximity preference

Notes

This latter percentage is developed by looking at the percentage given to broad classes of charities (religion, health, education, arts, etc.), and estimating the percentage that type of charity provides to poorer people. For example, an estimated 40 percent of all individual giving is to religious organizations, and religious organizations are estimated to give 20 percent of received gifts to the poor, so an additional 8 percent of giving flows to the poor indirectly through religious organizations.

This high estimate is consistent with a desire to provide an upper bound. Reich (2006: 30), who wishes to minimize his estimate of giving to the poor, cites an estimate that religious organizations provide only about 5 percent of received gifts to the poor. This would suggest only an additional 2 percent of giving flows to the poor indirectly through religious organizations. Similarly, the Indiana survey attributes money given to schools for scholarships (17 percent) as aid to poor, implying an additional indirect 3 percent of giving to poor (since 18 percent of all giving is to education). However, much of the scholarship and financial aid likely goes to the middle class and other non-poor.

This figure comes from Sansing and Yetman (2006), and appears to be based on 2000 tax year data.

Technically, they are called “private foundations” and “public charities” in the tax code, but we will refer to them simply as “foundations” and “charities.”

An extreme interpretation of this view is that Leona Helmsley's $8 billion endowment for “the care and welfare of dogs” (Madoff 2008) is no less praiseworthy than if she gave the money to people in Darfur.

This position strikes most people as extreme because it requires givers to impoverish themselves to the level of the poorest people. (Indeed, some philosophers regard Singer's conclusion as a reductio ad adsurdum for any thoroughly utilitarian approach.)

This follows from the fact that marginal utilities are higher for poorer people (that is, a poor person values a $1 gift much more than a wealthy person), and giving most to a poor person yields the greatest increase in utility.

For example, many philosophers, including John Rawls in A Theory of Justice (1971), argue that a simple maximizing strategy cannot adequately account for distributive justice. They argue it misses special duties we owe to the “least well off,” whose basic needs are unmet.

To see the intuition for η as a measure of inequality aversion, suppose η = 0 (no inequality aversion), then U = Income + Gift. Then total utility in Equation(1) is (1 + p_Near)(Income_Near + Gift_Near) + (Income_Far + Gift_Far), and all of the gift will go to the Near in Equation (1) so long as p_Near > 0. On the other hand, as η becomes larger, allocations will more and more favor the poorer beneficiary.

Arrow et al. (1996: 137–138) note: “Although all social welfare functions have been criticized for assuming interpersonal comparability of utility, there seems to be no way of addressing the ethical issues involved in making decisions affecting different generations without making some assumptions implicitly or explicitly about interpersonal comparability.”

Individuals and governments often debate issues of endowments. One such issue is that of social security taxation and benefits. In this context, a citizen's retirement plan is an endowment for benefiting his retirement years. And for this reason, the issue of the future versus the present, including its related aspect of affinity, has been an ongoing part of the social security debate in modern industrial societies.

To do this in our simulated solutions of Equation (2), we add a first step (not shown in Equation (2)), in which we estimate total interest in the future, and then divide the total gift with estimated interest between the present and future beneficiaries to maximize utility. For example, if we had $100 to divide between this year and next year, and interest was 5 percent, we would estimate the total interest, and then divide the total into two equal payments of $51.22. Note that this is not what utility maximization necessarily suggests: When there is interest, a dollar not given today is a dollar plus interest in the future, so greater utility can be achieved by saving more (paying out less) in the present. Rawls (1971: 286–287) uses this potential excessive savings as a critique of utilitarianism. The difference in savings occurs because interest is an opportunity cost of giving now, but our shorthand formula does not explicitly consider the countervailing benefits. For example, when we spend now on a student's education, that student's human capital and earning potential is increased. Therefore, there is more value in a scholarship today rather than having the student wait a year and receive the same scholarship with an additional 5 percent interest. In general, any person who faces constraints (e.g., lack of money for education or job training, high-interest debt, etc.) will value a gift today more than the gift in a year plus interest. So long as each year's gift is expected to create value at 5 percent or more per year, the right solution approach anticipates interest and yields equal payments when there is no time preference.

Note that these figures are consistent with the university numbers above. Returns are roughly 10 percent, and academic inflation is roughly 5 percent, for an after-inflation return of 5 percent. Payouts are around 5 percent. Also note that, as discussed in Swensen (2000), most universities have an explicit policy of “intergenerational equity” or zero time preference.

By the time of this writing in February 2009, global equity markets had fallen about 50 percent from February 2008. University endowments suffered along with the downturn, although the extent of decline had not been precisely disclosed (typically, exact results are not reported until the end of the fiscal year, and most universities have a 30 June 2009 fiscal year).

Note that Yale recently approved plans to increase the number of undergraduates from 1,300 to 1,500, which is the first increase in over 45 years. This suggests a per-year growth rate in beneficiaries of roughly 0.3 percent.

Uncertainty about the existence of future persons is also relevant to charitable allocations (Revesz 1999: 1001). If we are not sure that a future generation will exist, we should allocate more resources to earlier generations, which are more likely to use the resources.

Evidence in Core et al. (2006) suggests a negative correlation between endowment size and efficiency.

For example, if hedge fund and private managers are a major source of donations, endowments used as precautionary savings would avoid investing in hedge funds and private equity (Massy 2008).

Equality of persons is an argument for complete impartiality, but other philosophy offers support for the idea of some justified discounting of other persons. We may choose to help those more to whom we are closer by connections of community, generation, etc., than those to whom we are not (e.g. Williams 1985; Oldenquist 1982; Donaldson 1990). The intuition that I have a slightly stronger duty to my community, my fellow citizens, and members of my generation, than I have to far distant strangers and future persons living centuries from now is not necessarily irrational.