Donor Advised Fund Policies and Intergenerational Justice

1 Introduction

Every year more and more money is contributed to donor advised funds (DAFs). These are charitable giving accounts hosted by a “sponsoring” nonprofit organization. To establish a DAF, a donor makes a tax-deductible contribution into the account; the sponsor gives the donor “advisory privileges” to recommend how the money in the account may be used. The donor “advises” the sponsor where to grant that money, such that the grant may happen (much) later than the tax deduction on contribution. Technically, the sponsor retains legal control over the money, but generally donor advice is followed. Meanwhile, the funds in the account are invested, growing tax-free, and the sponsor organization charges management fees. Donors, also known as “donor advisors”, benefit from using a DAF because they get an immediate tax deduction (often for appreciated assets) and they get time to decide where to grant the funds. These benefits and others have made DAFs extremely popular, with over two million DAF accounts and donations to DAFs more than doubling in the past five years from $36.7B in 2018 to $85.5B in 2022 (National Philanthropic Trust 2023).

This growth raises concerns about how long that money stays in DAFs before being redistributed. The DAF Research Collaborative (2024) calculated the shelf-life of DAF money, showing that more than half of DAFs distributed between half and all the contributions within three years. However, some DAFs hold onto the money for much longer. The questions this raises about timing include “do DAFs delay for too long the time before a donation is actively put to use in benefiting society?” (Colinvaux 2022; Galle 2016); and “do DAFs provide decision-making power (in practice, if not in law) for too long a period of time to donors by way of advisory privileges?” (see, eg, Colinvaux 2022). DAF sponsors exercise some control over the timing of DAF grantmaking through internal policies. Three specific types of DAF policies deal directly with the timing of the distribution of philanthropic resources: inactive account policies, endowed DAF policies, and succession plan policies. We investigate how these DAF policies affect the timing of distribution of DAF resources and whether this reflects what is fair, or just, for society.

Murray (2023) proposed the use of normative theories of intergenerational justice to address questions about the delay in timing that DAFs create as intermediary vehicles. These theories are a set of political theories focused on articulating the obligations owed by the current generation to past and future generations (Gosseries and Meyer 2009). They can be used to consider the obligations that current generations have to save resources or preserve decision-making freedom for future generations, as well as what obligations apply once sufficient resources have been set aside. In considering different mechanisms for applying intergenerational justice theories to the DAF context, Murray (2023) focused on the duties that apply to directors/trustees of the public charities that sponsor DAFs. He drew specific attention to the role of DAF sponsor policies in this process (Murray 2023, 64, 66). In this article, we build upon this work by exploring specific DAF sponsor policies that relate to timing. We ask the question, to what extent do these policies reflect principles of intergenerational justice? This involves checking consistency with norms of intergenerational justice, rather than necessarily requiring the intentional application of those norms in the policies.

We analyze the inactive account, endowed DAF and succession plan policies from the largest DAF sponsors in the US. The empirical findings from this analysis provide an opportunity to explore whether principles of intergenerational justice are already incorporated within DAF sponsor policies. If so, this may partly answer commentators’ concerns about undue delay. To the extent that policies do not demonstrate application of intergenerational justice, there might be scope for amendments.

We provide evidence that DAF policies create boundaries in which DAF grantmaking may reflect the norms of intergenerational justice, especially in saving philanthropic resources for the future, but not necessarily in setting sufficient emphasis on distributions to the current generation or responding to changed circumstances. Inactive account policies establish minimum expected levels of distribution and thus mandate some distributions but are fairly permissive in terms of enabling material saving for the future. Endowed DAF policies set out maximum permitted levels of distribution encouraging saving for the future to an even greater extent. Succession plan policies regulate the extent of accumulation by substituting a new regime when the initial donor advisor dies or is incapacitated. These rules might require the immediate distribution of any remaining funds in the DAF account or support their ongoing retention. Succession plan rules and the DAF sponsor variance power which overrides the standard endowed DAF rules, do a good job of providing future generations with broadly equal decision-making power over philanthropic assets. Based on our review of these three sets of policies, we offer recommendations for DAF sponsors to further integrate norms of intergenerational justice in their policies.

2 Theoretical Framework

DAFs exist within a much broader set of social and political institutions that jointly help (or hinder) social cooperation and the existence of a political society over time. Determining whether DAFs are good for society, or whether they unduly delay the application of donations or provide donors with too much power, requires some understanding of broader institutional settings, such as the level of redistribution achieved by the tax and transfer system, the impact of government policies on future generations, and the degree of decision-making power over social and cultural programs already accorded to those wealthy enough to be DAF donors. In the United States context, these broader settings are mildly redistributive (Causa and Hermansen 2017). There is an emphasis on current rather than future generations (Reich 2018, ch 5), and while representative democracy ensures a variety of interests are considered in decision-making about social and cultural programs, wealthy elites have a disproportionate influence (Bartels 2016; Saunders-Hastings 2022, 46–52).

Much of the DAF literature discusses current or proposed tax rules, typically without a clear normative benchmark (Colinvaux 2022; Galle and Madoff 2023). The closest that the literature comes to this is in analyzing DAFs as an exception to a “grand bargain” reached between wealthy philanthropists and the rest of society when the tax rules were set for private foundations (Reiser and Dean 2023: 6–11). The contours of that grand bargain permitted wealthy individuals to influence social and cultural communal projects via private foundations, but imposed “timing, targeting [to restrict private benefits] and transparency requirements” (Reiser and Dean 2023: 7–8). An outcome of political power, rather than an application of normative standards, the “timing” conditions involved the imposition of a 6 %, now 5 %, annual payout rate, but fundamental to the choice of rate was that it was selected to permit the perpetual existence of private foundations (Reiser and Dean 2023: 8, 121–122). Thus material benefits must be distributed to the current generation, but not so as to preclude transferring benefits to future generations. DAFs, though, exist within public charities, not private foundations, such that they are not subject to this grand bargain, reflecting a “lacuna” (Berman 2020). Further, talk of materially increased taxes on university endowments and removal of tax exempt status from various classes of nonprofits suggests some weakening of grand bargains reached in the nonprofit space.

However, there is a growing body of work that seeks to explicitly engage normative standards (in relation to private foundations and DAFs) and to consider how those standards might be linked to fiduciary duties and the role of these duties in generating rules for DAFs under DAF sponsor policies (Murray 2023), along with the impact of those policies (Heist and Stone 2023). This builds on application of particular interpretations of intergenerational justice by commentators in other contexts such as private foundations (Cordelli and Reich 2016).

Intergenerational justice is a set of political theories focused on the obligations owed by the current generation to past and future generations (Gosseries and Meyer 2009). These obligations can be interpreted in terms of the distribution of resources between generations so as to satisfy basic social and economic needs (Birnbacher 2006: 34; Gaspart and Gosseries 2007) and hence can directly help answer questions about whether DAFs unduly delay the provision of benefits. Further, the obligations owed to past and future generations can be conceived in terms of the distribution of decision-making freedom (Murray 2021, Ch 8; Thompson 2009). Accordingly, intergenerational justice can also provide standards for the duration of respect that ought to be accorded to donor advisory privileges. For instance, this sufficientarian standard has been interpreted as meaning that future generations should be given material choice over distribution decisions, subject to some restrictions based on recognition of DAF sponsors as intergenerational entities, that might require, for instance, that the assets remain dedicated to charity and that a period of time of protection (for instance, that of a generational cohort of 20–30 years) be accorded to initial donor choices (Murray 2021, Ch 8).

What precisely does intergenerational justice require? It is worth reiterating that intergenerational justice consists of a range of different political theories that seek to articulate intergenerational obligations from various political perspectives: liberalism, communitarianism, etc. Liberal theorists have framed these obligations in prioritarian (Meyer and Roser 2009, 234–5; Adler and Treich 2015, 294–7), sufficientarian (Laslett 1992, 29–30, 44–45; Meyer 2021), utilitarian (Birnbacher 2006, 32–33), and libertarian (Fleischer 2023) terms. In the context of DAFs, Murray (2023) suggests focussing on a “sufficientarian” interpretation because this better fits the institutional role of DAFs in maintaining social capital and supporting long-term and innovative civil society approaches.

A sufficientarian approach would mean that the current generation should not impose costs on future generations that result in the world being handed on in a state that fails to meet a minimum threshold standard for members of future generations. The threshold can be defined in different ways. However, one common way is to do so in terms of a level of resources to meet basic needs (Meyer 2021), which, while less commonly noted, could include sufficient resources to support the institutions of civil society (Murray 2023). If the threshold is low and relates to fundamental needs, then the current generation would not be required to materially sacrifice their own wellbeing by saving to benefit future generations. Indeed, once the threshold has been reached, it is possible to switch to prioritarian principles for allocating the remaining resources amongst the current generation (Shields 2016, ch 2; Meyer 2021). Prioritarianism focuses on redistributing resources to the most disadvantaged (giving them a priority) but does not require equal division of resources to the extent that socio-economic differences improve the absolute position of the most disadvantaged, for instance, because they incentivize greater productivity (Rawls 1999, 57–73).

It is also possible to view the minimum threshold in terms of the degree of decision-making freedom over how to distribute philanthropic funds in support of civil society. Murray (2023; 2021, 237–240) has suggested that this threshold should typically be a level of decision-making freedom for future generations that is at least equal to that of the current generation, albeit future generations might be expected to show some respect to donor wishes as to what civil society projects to support, provided this does not materially detract from decision-making freedom. This does not suggest that a particular future generation should have the power to immediately distribute all assets held by DAFs, on a whim. Just as for current generations, the decision-making of future generations as to the distribution of DAF resources, should occur in accordance with the principles of intergenerational justice discussed above, with sufficientarian principles acknowledging a role for establishing and maintaining a pool of resources to support civil society. Indeed, as many DAFs are intended to last for longer than a generation (potentially in perpetuity), their role in supporting civil society across these intergenerational timespans itself plays a part in expanding the decision-making freedom of future generations, by expanding the range of choices available.

A sufficientarian approach to DAFs asks the question: “how much philanthropic funding should be left to the next generation (through DAFs) in order to meet the basic needs of civil society in the future?”. Once that threshold is met, the remaining DAF resources should be used in a prioritarian way to address the most pressing needs – typically, as we suggest below, of the current generation. We have termed this a sufficientarian/prioritarian approach. The challenge to answering the question as it relates to DAFs and, more broadly, to all philanthropic organizations, is the pluralistic nature of the philanthropic sector. The basic needs of current and future generations will change over time, and some types of need – for instance health or spiritual needs – might vary in quite different ways over time, with some needs increasing and others decreasing. Moreover, each DAF sponsor may have a mission and values that prioritize certain types of needs and that moderate their approach. Application of intergenerational justice principles will therefore result in quite different expectations, depending on the context.

For example, consider what it means to adopt a sufficientarian approach (with prioritarian principles applied once the threshold is reached) for a national DAF sponsor which has a very broad range of charitable purposes and hence a broad range of supported civil society organizations. The DAF sponsor would be expected to assess the degree to which the basic needs (connected with that range of charitable purposes) of current and future generations are expected to be met. This would include considering the ways those needs might change over time, such that their funds should be set aside to assist civil society bodies to develop responses to those changing needs or to better meet existing needs. The analysis should include some understanding of influences external to the DAF sponsor. For instance, the impact of state social welfare and environmental policies (amongst others) and wider systems such as the tax and transfer system. Undoubtedly, views will differ about the likely effects of state policies and systems, especially into the future. However, it seems plausible to assume that changes to state policies and systems will not have a dramatic effect on the needs associated with the range of charitable purposes as a whole, albeit those changes may affect specific purposes. Absent a cataclysm, it also seems plausible to assume that basic needs will remain relatively stable over time. If so, then assuming continued per capita GDP increases and material levels of ongoing donations, this would suggest that material levels of new DAF contributions should be spent over the short-to-medium-term.

For DAF sponsors with a more focused charitable purpose, the exercise is likely to differ. For instance, for community foundations with a particular geographic community focus and single-issue DAF sponsors (such as university DAFs), the DAF sponsors should have a sense of external factors such as applicable government policies and their potential impacts, as well as an understanding of existing and likely future levels of need for services. However, the expected understanding would be more focused on the particular purpose or geographic area, and these DAF sponsors seem likely to have a deeper understanding in these focussed areas than the all-encompassing national DAF sponsors. There is also greater potential for variability in matters such as the level of the more focused needs and donation levels over time. In the context of a religiously-affiliated DAF sponsor, if a religious organization expects its future growth in membership to be in developing countries where the needs will be greater than the contributions provided by its membership in developed countries, then that organization may encourage DAF sponsors to leave funds for future generations. Likewise, a community foundation in a community with a shrinking economy might encourage donors to leave DAF resources to the community foundation or successors that will be able to help address future needs of the community.

DAF sponsor policies are then a tool by which DAF sponsor officers and directors can implement the position reached by application of these intergenerational justice principles; and seek to influence individual DAF donors to make more rapid distributions to benefit current generations, or to curtail the duration of donor decision-making by limiting the extent of time for which advisory privileges run. DAF sponsors have a range of policies relating to issues such as opening a DAF account, making contributions and associated income tax requirements, investment, fees for services, grantmaking, succession and family involvement, privacy, and philanthropic services for DAF advisers. As set out in the Introduction, we seek to answer the question: to what extent do current DAF sponsor policies reflect principles of intergenerational justice? This more general question can be expressed more specifically as follows and it is the grantmaking and succession rules that are most pertinent to these specific questions relating to timing:

1. Do DAF sponsor policies permit but also restrict the accumulation of funds to enable saving for future generations? Do DAF sponsor policies compel (and have the potential to change that level of compulsion) distribution to benefit the current generation? Policies with such functions could enable saving to meet a sufficientarian threshold of need for future generations, but also ensure that once that threshold has been met, payouts can be compelled to the current generation to satisfy prioritarian requirements. There are two primary such sets of rules in DAF sponsor policies. First, “inactive account policies”, that set out minimum expected levels of distribution and thus permitted saving over a period of time. Second, endowed DAF policies which set out maximum permitted levels of distribution and hence permitted saving over time. Finally, there is a residual set of rules that regulates the extent of accumulation; succession plan rules that substitute a new regime when the initial donor advisor dies or is incapacitated. These rules might require the immediate distribution of any remaining funds in the DAF account, or enable their perpetual retention. There are different ways that inactive account policies, endowed DAF policies and succession plan rules might enable saving and compel distributions. For instance, these policies might require payouts from all DAF accounts each year, e.g. a 5 % payout as is the case for private foundations. Alternatively, policies might restrict distributions and mandate a level of savings, as discussed for endowed DAF policies. Rather than mandating particular levels of distributions or savings, policies might instead seek to encourage donor behavior, by encouraging donor advisors to think about distributions and savings on a regular basis, for instance by providing mechanisms for identifying and then contacting donors who do not propose grantmaking over a set period of time. Another encouragement strategy might be for policies to set out a DAF sponsor preference for the balance between current and future generations and then exhort donor advisors to consider this in their grantmaking. While the pluralistic nature of the philanthropic sector makes it hard to be specific about how we might expect these approaches to be reflected, as noted above, we probably do expect the application of sufficientarian/prioritarian principles to a national DAF sponsor to require or encourage the majority of new contributions and investment income to be distributed over the short-to-medium term, perhaps 20–30 years to reflect a generational cohort.

1. As intergenerational justice also requires a fair distribution of decision-making power between current and future generations, a further question is whether DAF sponsor policies enable future generations to make choices about the level of distribution and saving and the identity of distribution recipients? Consistent with sufficientarian principles, we might expect all types of DAF sponsors to enable material decision-making on these matters by future generations, subject to the restriction that the assets remain dedicated to the DAF sponsor’s charitable purposes and potentially, to show respect to donor choices, with something like a 20 to 30-year period of greater protection for initial donor choices. Succession plan policies are the primary set of rules that articulate who (and hence the members of which generation) can make such decisions. However, we also briefly consider the interaction of succession plans with endowed DAF policies, since endowed DAF policies are a potential means of restricting the level of decision-making power of future generations if they can apply over multiple generations.

3 Data and Methods

This study analyzes policy documents from the largest 158 DAF sponsoring organizations in the US, based on 2019 year-end DAF assets from the IRS Form 990, Schedule D, Part I, line 4. The researchers initially sought to collect the 150 organizations to ensure a large enough sample for each policy type; ten more organizations were added because of challenges with data collection. Two organizations were dropped because of apparent flaws in the 990 data (sometimes nonprofits erroneously file Schedule D). The final sample included national sponsors, community foundations, religiously-affiliated organizations, and a few universities (see Table 1). The specific policy documents collected were: inactive account policies, endowed DAF policies and succession plan policies. Policy information was primarily collected in the form of content published on a web page, PDFs published online, or a PDF emailed to the team. The research team used advanced Google searches to search the websites of each organization for policy documents. If policy documents were not found online, our research team contacted the organization via email or telephone. Some organizations responded that they were not willing to share their policy with us. These were coded as “withheld” in the table below and only pertained to inactive account and endowed DAF policies. If an organization did not respond within a two-week period from the first email, a second email was sent, and phone calls made. If no response was received after multiple attempts, the specific policies for the organization were considered “missing”. Some types of policies were more readily available than others, which is why the number of “missing” observations varies by policy-type. Therefore, while our total sample of organizations was 158, the different types of policies had differing numbers of observations.

Of the 158 organizations in the sampling frame, data on inactive policies was obtained from 103 organizations, data on endowed DAF policies from 144 organizations, and data on succession plans from 121 organizations. Table 1 describes the full sample by sponsor type, and the available data collected for each policy type. A “Yes” means we collected data on that policy and the organization does allow that option. A “No” means we collected data on that policy and the organization does not provide that option. For example, an organization may have a policy on succession plans but does not provide an option for naming charitable beneficiaries. Missing data is dropped from the analysis of each category. While we do not assume that policy information is missing at random, an in-depth analysis of the reasons for missing policy information is beyond the scope of this study. Appendix A shows the percentages of missing policies by sponsor type. A Chi-square test was used to determine if the proportion of missing policies was significantly larger than what would be expected for each sponsor type. A chi-square statistic of χ² = 3.63, with 3 degrees of freedom and a p-value of 0.30 indicates there is not enough evidence to substantiate a relationship between the type of sponsor and whether they provided these policies.

Policy documents were analyzed using qualitative analysis software MaxQDA to code text that included information about any of the policies related to this study. Two authors worked on the qualitative analysis and reached an intercoder reliability rating of over 90 %. We do not cite DAF sponsor names.

4 Inactive Account Policies

Inactive account policies apply to DAF accounts from which no or insufficient grantmaking has occurred over a period of time. While DAF sponsor policies vary in relation to the definition of “inactivity”, the relevant time-period and responses to inactivity, the vast majority of DAF sponsors (83 %) had written inactive account policies. Figure 1 shows the typical stages exhibited under inactive account policies. This is consistent with the California Attorney General’s office finding that over 95 % of larger California-based DAF sponsors had inactive account policies (California Department of Justice 2022). Further analysis, found in Appendix B, indicates that those with inactive account policies tended to be the larger DAF sponsors than those who did not have a policy or those who withheld their policy. These differences were not statistically significant mainly because of the relatively few organizations that did not have a policy or withheld their policy.

Figure 1:

Pattern of inactive account policies.

First, the DAF sponsor identifies when an account is considered “inactive” due to an absence of grantmaking for a defined time period (labeled T₁ in Figure 1). For example, one policy states, “If no account holders recommend grants during a period of 30 months, the account will be considered inactive.” The average period for T₁ was 36 months. Second, the DAF sponsor attempts to contact the donor advisors to encourage grantmaking: “Staff will contact the donor advisor(s), to make them aware of the low grantmaking activity, and discuss the donor’s philanthropic goals and grantmaking intentions for activating the grantmaking of their fund.” Third, donors are permitted time to recommend a grant or take some other action (labeled T₂). The average time for T₂ was 18 months. Fourth, if a donor does not act, the DAF sponsor either starts issuing grants from the account on its own initiative, potentially with regard to the DAF’s history of grantmaking or a donor succession plan or closes the account and redistributes the assets.

As explained above, sufficientarian principles would countenance a degree of retention of funds within DAF accounts to meet the needs of future generations and support civil society, funding social capital. However, absent material increases in government support for health, education, etc., trends such as increasing GDP and massive growth in DAF contributions suggest that future generations are likely to be better placed to meet their basic needs than are members of the present generation. Hence, material levels of DAF contributions should be distributed fairly quickly. Inactive account policies are clearly a nod in this direction, demonstrating an organization’s preference for current grantmaking.

In many cases, however, inactive account policies had vague language around some aspects of inactivity:

The [Community Foundation] periodically checks in on donor-advised funds that are not being used to ensure the [Community Foundation] understands Fund [donor] Advisors’ plans for their funds.

The policy neither specifies a clear definition for inactivity nor the time in which the sponsor will act, nor the conditions for resolving the inactive account. In almost all cases, distributing a grant qualified an account as being “active”, however policies either did not specify a particular minimum amount, or only a nominal amount.

Additionally, it was not uncommon (26 % of policies) to provide donor advisors with some flexibility in explaining away a lack of grantmaking by reference to a growth or long-term charitable project plan, although in most – but not all – cases the policy set out a maximum time limit for such planned growth, with three years being the maximum. Some policies (11 %) also permitted a broader range of explanations, generally linked to life events that affected the donor, like severe illness. It is unclear from the policies whether – and to what extent – DAF sponsors provide donors with assistance in developing long-term charitable project plans. That policies with a time limit extended only to a three-year maximum is broadly consistent with the suggestion above that sufficientarian principles would support only modest savings for the future.

Some policies also permitted a long period of time before the DAF sponsor implemented a fourth stage response, with an 11-year period being the longest permitted period in the sample. However, a maximum inactivity period of 11 years seems modest in intergenerational terms with something like a 20 to 30-year time period more reflective of a generational cohort. More concerning is the lack of detail in many policies about repeat periods of inactivity, interspersed by minimal grants, which would potentially permit a far longer deferral of grantmaking. A very small minority of policies deliberately address this issue:

Dormant Funds are those with no grant activity in the previous three or more calendar years. In the event of repeated or sustained dormancy, the Trust may remove all advisors and operate the fund in a manner consistent with other permanent funds…

The above presents a possible approach for rendering inactive account policies more meaningful. However, it could potentially be matched with a requirement for grants (that reactivate a DAF account) to be based on a proportion of the funds in the DAF account or the income earned. For instance, several policies required grants of at least “5 % of their net assets” to qualify for reactivation.

There are, however, two further approaches adopted in many DAF sponsor policies that bolster the inactive account policies and, where present, go some way to addressing concerns about the vagueness of the inactive account policies. First, more than half (55 %) of DAF sponsor policies dealing with inactive accounts contained a general statement of norms emphasizing that DAF accounts must be employed to achieve the DAF sponsor’s purpose, involving grants to benefit the community now, rather than being accumulated for the future. For instance:

It is the general policy of the Committee that the assets of Philanthropic Donor-Advised Funds actively help the community, rather than accumulate.

This is potential evidence that DAF sponsor directors are giving consideration to matters of intergenerational justice in formulating fund policies since it highlights the importance of benefitting the current generation. While such aspirational statements do not lay down rules, they may play an educational and motivational role for DAF sponsor officers and donor advisors (as to the motivational role of statements of purpose, see Langford 2020: 958).

Second, 13 % of DAF sponsors’ inactive account policies (including some national DAF sponsors) included a sponsor organization-level minimum distribution requirement. In each case, the minimum distribution level was specified as 5 % of the DAF sponsor’s average net assets over a five-year period, which reflects – in broad terms – the payout requirement for private foundations (Fishman et al 2021: 794–803). For instance:

[DAF sponsor] expects that in the aggregate its grant distributions will exceed 5% of its average net assets on a fiscal five-year rolling basis. In order to achieve this level of grant activity, [we] will identify donor accounts from which grants over any five-year period totaled less than 5% of such account’s average assets and will contact the account donors to request that they recommend grants of at least 5% of the account’s average assets over the five-year period. If a donor does not provide grant

recommendations within 60 days of such a request, [DAF sponsor] reserves the right to transfer an amount up to 5% of the account’s average assets over the five-year period from the donor’s account to the [DAF sponsor general giving fund].

There remains a potential for DAF grants to be made to other philanthropic intermediaries (such as other DAF sponsors) rather than operating charities (Murray 2020, 271–2, 294; Brakman Reiser and Dean 2023, 69–70). However, organizational-level policies on minimum payout rates indicate that DAF sponsor directors and trustees are aware that their fiduciary duties require informed decision-making and responsible consideration in relation to the amount of distributions each year and the designations of those distributions (as to the duties, see Murray 2023, 62–4). This interpretation is bolstered by the fact that 73 % of the inactive account policies that included organization-level payout requirements also incorporated statements of general norms relating to active grantmaking to achieve the DAF sponsor’s charitable purpose, of the type discussed above. Organization-level requirements are also a reminder that the decision-making authority delegated to donor advisors (via advisory privileges) is limited and that, ultimately, the DAF sponsor directors or trustees are responsible for oversight and distribution decisions (see Murray 2023, pp. 62–4).

An organization-level minimum distribution requirement would likely result in a DAF sponsor retaining enough investment income to maintain the real value of donations and possibly to grow at a very slow rate (for studies on the effect of the 5 % payout rate for US private foundations, see Toepler 2004; Leat 2016: 86–8). Such an approach is relatively neutral between generations and so is potentially consistent with sufficientarian principles (Murray 2021: 213–215) in that it enables some modest saving for future generations. However, it is rough and ready and does not take account of matters such as whether the needs served by a DAF sponsor are likely to change over time (e.g., a religious DAF sponsor with a religious community that is about to materially increase in size or a community foundation sponsor focussed on an underserved community that is increasing) or whether donations are likely to spike and then drop again (e.g., as part of the wealth transfer from the baby boomer generation). To reflect changing thresholds or levels of need of the current generation versus the future generation, inactive account policies should ideally respond to such changes over time, for instance by permitting variation in the required payout rate.

While evidence from the private foundation context suggests that a minimum rate can lead to industry norms treating that minimum rate as a maximum (Deep and Frumkin 2005) aspirational statements of active grantmaking norms might nevertheless encourage DAF sponsors and donor advisors to consider giving above 5 %. Further, at the end of the day, the 5 % minimum is a default or fallback position, rather than a mandatory requirement. DAF sponsors can still distribute at higher percentages. Thus, the DAF sponsor board may choose to distribute more, or less, in a given year – whether by availing itself of the flexible wording used in policies or by amending the policy. The important thing is that in amending the default position (if it chooses to do so), the board will be obliged to consider how DAF sponsor assets ought to be distributed between generations, in relation to which intergenerational justice principles will be critical.

Nevertheless, only a low percentage of DAF sponsors have established organization-level distribution requirements, so we may question the extent to which directors or trustees of the remaining sponsor organizations have engaged with intergenerational justice principles or actively considered what an appropriate level of distributions might look like. The low take-up of default organization-level payout requirements represents an opportunity to bring more rigor to inactive account policies and, at the same time, to encourage board level conversations about the importance of intergenerational thinking in potentially perpetual DAF sponsors.

5 Endowed DAF Policies

While the vast majority of DAF sponsors have inactive account policies that encourage a minimum level of distributions each year, less than half of sponsors (44 %) also included in their policies the option for an endowed DAF (see Table 1). Endowed DAF policies have three key features:

1. The level of grants permitted is capped at a level intended to permit the DAF to exist in perpetuity – by setting a cap that retains the real value of the DAF.

2. Donor advisory privileges remain in place, such that donors and other advisory privilege holders can continue to recommend their preferred grant recipients each year.

3. The endowed DAF policies apply until a succession plan is triggered or, potentially, intervention occurs under inactive account policies. In effect, this means that the endowed DAF policies typically apply during the initial donor advisors’ lifetimes or their incapacity (initial donors might, for instance, appoint spouses or other family members as initial rather than successor advisors). Sometimes succession options can continue to apply until the death or incapacity of successor advisors, in which case successor advisors may be limited in their ability to remove the grantmaking cap. When succession options are then activated upon death or incapacity, this potentially opens up access to endowed funds. We discuss succession plans below.

The cap is typically set by reference to a DAF sponsor’s spending policy, which provides a process for the DAF sponsor to annually determine a rate based on industry norms and state laws applying to endowments. In most instances, the rate was expressed as a percentage (e.g., 5 %) but subject to variation each year by the DAF sponsor. The average rate for DAF sponsors in the sample was 4.4 %, with a range of 4 %–6 %. Some policies did not specify a percentage but referred more generally to limiting distributions to income and not principal.

One of the reasons that DAF sponsors provide endowed DAF policies is because they may wish to grow their total assets under management. Organizations that offered endowed DAFs were found to be smaller, based on 2019 year-end DAF assets, compared to those that did not offer endowed DAFs (see Appendix B). The hypothetical examples provided in the literature review, for instance, of a religiously-affiliated DAF sponsor that expects future religious membership growth in developing countries, or a community foundation that anticipates a shrinking local economy, demonstrate why a DAF sponsor might wish to build up resources. As will be explained in the next section, many succession options that apply on donor advisors’ deaths include allocation of DAF funds to the DAF sponsor’s discretionary giving fund or general endowment fund. Endowed DAFs are essentially deferred gifts to the DAF sponsor, over which the donors have some advisory privileges until their death.

As discussed under Inactive Account Policies in relation to organization-level payout rates, a payout rate around 5% – or a little lower for endowed DAFs – is likely to mean that the endowed DAFs maintain their real value over time, with the possibility of some slow growth. This ensures saving for future generations and is potentially consistent with some applications of sufficientarian principles in that a modest level of support for civil society is set aside for future generations. However, the endowed DAF rate will have further effects. That is because while the organization-level payout rate serves as a default floor, the endowed DAF rate is a ceiling. It therefore limits a DAF sponsor’s ability to respond to sudden and significant changes in need that shift the current generation below the sufficientarian threshold (requiring immediate distribution of resources to the current generation so that they can meet their basic needs). For instance, the Great Depression of the 1930s caused significant need for basics such as food and shelter for a material proportion of the US population (see, e.g., Boyer 2001, 184). Further, material events might also affect our specification of the sufficientarian threshold itself. COVID-19 provides an example of an event that has arguably reduced thresholds relating to health needs, potentially leading to less need to save for the future. Prior to COVID-19, we might have conceived of a sufficientarian threshold in terms of the level of resources needed to ensure a certain base level of health care for all. Post COVID-19 we might shift that threshold downward so as to require resources only for health care needed to avoid serious illness or death, with the threshold slowly increasing to cover a broader range of health needs over time (analogously with Rawls sufficientarian “just savings principle” which was intended to slowly build resources over time for establishing just institutions, thus slowly increasing the threshold over time: Rawls 1999, 251–8). In each of these such situations, our sufficientarian/prioritarian approach would require DAF sponsor policies to have mechanisms that respond to the situations by adjusting the weighting of distributions in favor of the present generation and away from future generations, meaning that there should be a mechanism to adjust the endowed DAF rate upward.

A ceiling on distributions would also preclude shifting to a prioritarian approach if we consider that the threshold has been reached (that is, enough has been set aside for the future) and that those alive now are in greater need than those in the future. In other words, once a modest level of support has been put aside for the future, DAF sponsor policies should encourage distributions to the present generation. The need for this approach is highlighted by the potential for further donations, with ongoing contributions to a DAF being a strategy adopted by a material proportion of donors. Indeed, research on community foundation and religiously-affiliated DAF sponsor organizations found that, over a four-year period, 68 % of DAFs received additional contributions (Vance-McMullen and Heist 2022,16). At a whole-of-sector level, the 2022 report from National Philanthropic Trust (2022, 15–16) shows that total contributions to DAF sponsors are about 150 % of grants and that total contributions have exceeded total grants in each year from 2017 to 2021. As discussed in the Theoretical Framework section, reaching the sufficientarian threshold might be anticipated to occur with some regularity for national DAF sponsors with a broad range of charitable purposes, while more variability in the threshold (and hence in whether a ceiling on distributions is problematic) might be expected for community foundations, single issue, and religiously-affiliated DAF sponsors. That might explain why most national DAF sponsors in the survey did not offer endowed DAFs, while endowed DAF policies were more prevalent for community foundations and religiously-affiliated DAF sponsors (Table 2).

The DAF policies reviewed do, however, contain a potential adjustment mechanism that could respond to the above concerns. That is, all policies contained explicit statements that the level of distributions is ultimately up to the DAF sponsor. In the sample, this typically occurred by way of a “variance power”. For instance:

“All contributions to the [DAF sponsor] are subject to the [DAF sponsor]’s variance power, as stated in Article [X] of its Articles of Organization, which gives the Board of Directors of the [DAF sponsor] the right to redirect the use of any fund under certain conditions to better meet the changing needs of the community.”

In contrast with standalone endowed charitable trusts, where the trustee has very limited ability to distribute the trust capital (Fremont-Smith 2004: 136–7; American Law Institute 2021: §2.04), the variance power clearly permits a DAF sponsor to increase grants above the cap in response to shocks such as COVID-19. Nevertheless, as the language of the example demonstrates, the variance power is a reserve power. It provides no policy or set of procedures for the DAF sponsor to consider whether the needs of the community have changed. Accordingly, the impact of the variance power on grantmaking is likely to depend on the general practice of the relevant DAF sponsor and on the policies and procedures relating to governance and decision-making of the DAF sponsor.

In relation to the distribution of decision-making power between generations, endowed DAF policies (in the same way as spendable DAF policies) still leave future generations with material decision-making power, in support of meeting a sufficientarian threshold of decision-making power to the next generation that is at least equal to the previous generation. That is for two reasons: first, as just discussed, the DAF sponsor variance power enables future generations of DAF sponsor decision-makers to override endowed DAF grant caps. Indeed, the variance power would facilitate the DAF sponsor board or trustees amending the DAF sponsor’s policies. Second, endowed DAF policies apply only until a succession plan is triggered or, potentially, intervention occurs under inactive account policies. As noted at the outset to this part, and explained further in the next section, DAF sponsor policies explicitly provide for choices around succession options linked to the death or incapacity of the initial donor advisors. It is true that those choices often include methods for extending the endowment cap, such as providing for successor advisors to the endowed DAF for a period or selecting a legacy endowment option. However, those are not usually the only options. Thus, endowed DAF policies actively encourage re-consideration of whether a spending cap is appropriate and usually provide a mechanism permitting greater control to be passed over to future generations. Further, successor advisors’ rights are often limited in different ways, with some successor advisors retaining some ability to terminate and distribute a DAF. This means that endowed DAFs restrict future generations’ decision-making ability far less than endowed charitable trusts.

6 Succession Plans

Most DAF sponsors for which policies were obtained included rules in those policies setting out what would happen to DAF assets when the donor advisor(s) passed away or became incapacitated: “succession plans”. Donor advisors were provided a choice of succession plan options and many DAF sponsors required that this choice be exercised and that a succession plan be in place at the time a DAF is established. The succession plan options fall into four categories:

1. Appointment of successor advisors to take on some or all advisory privileges for either a period of time or indefinitely.

2. Conversion of the DAF account into a permanently endowed fund (‘legacy endowment’).

3. The internal reallocation of funds within the DAF sponsor itself by way of ‘grant’ to a discretionary giving fund or general endowment fund.

4. Distribution of the DAF account by way of grant to charitable beneficiaries.

Sometimes policies about succession plans included vague language about the options, such as “Advisor(s) may request that portions of the fund be administered in a variety of ways upon the death or incapacity of the fund’s last surviving advisor.” The statistics presented in Table 1 about each type of option only include policies with clear evidence for a particular option or the lack of that option. Missing data on succession plan options ranged from 25 % (for successor advisors) to 36 % (for charitable beneficiaries). Therefore, it is possible that more DAF sponsors offer the various options than we report. They just don’t include this information in their available policies. An analysis of missing data on succession plans found that smaller organizations were more likely to be missing available data about succession plans, (see Appendix B).

From an intergenerational justice perspective, policies on succession plans are particularly relevant to the distribution of decision-making freedom between generations.

7 Discussion & Recommendations

DAF sponsor policies reflect intergenerational justice principles to some extent. In particular, the widespread adoption of inactive account policies, backed up by broader policy statements of norms of active grantmaking are consistent with an understanding that resources need only be accumulated for future generations up to a modest limit (the sufficientarian threshold), beyond which prioritarian principles might often be expected to demand a focus on the current generation by way of immediate distributions to civil society organizations. The policies could nevertheless be improved by better dealing with repeat periods of inactivity and by adopting a definition of activity based on a percentage of a DAF account balance rather than a nominal dollar amount. Further, succession plan policies that provide for legacy endowments or place spending caps on successor advisors support a build-up of funds, which also seems contrary to the general need to favor relatively quick distribution. The intergenerational justice argument for immediate distribution is strongest for national DAF sponsors with very broad charitable purposes and it is consistent with these expectations that national DAF sponsor policies were less likely to offer an endowed DAF option (and that national DAF sponsors included some prominent examples of organization-level payout rules). DAF sponsors with more specific charitable missions, such as community foundations and religiously-affiliated organizations, may have more justifiable reasons for preserving charitable resources for future generations (that is, selecting an increased sufficientarian threshold for those future generations).

However, there is significant room to improve the way in which policies deal with changed circumstances. A minority of DAF sponsor policies contained sponsor-wide minimum distribution requirements. While these minima likely act to prompt consideration about what the right level of distributions is each year, the danger is that DAF sponsors simply look to the stated minimum percentage, rather than thinking about whether that percentage should change due to the changed circumstances. The policies could be improved by explicitly requiring (reminding) DAF sponsors that they need to consider changed needs of those supported by the charities to which they distribute. A similar amendment could be made to endowed DAF and legacy endowment policies to the extent that those policies provide a process for the DAF sponsor to annually determine an appropriate rate of distribution, which could be tailored to also take account of changed circumstances.

DAF sponsor policies appear to generally maintain a material level of decision-making freedom for each generation, consistently with the demands of intergenerational justice. Indeed, the time limit on many endowed DAF policies actively encourages re-consideration of whether a spending cap is appropriate when a succession plan is triggered, while many of the succession plan options themselves provide mechanisms permitting greater control to be passed over to future generations. Some divergence from a sufficientarian threshold of equal decision-making power for each generation occurs in relation to the succession options that enable legacy endowments or that enable the transfer of DAF accounts to a DAF sponsor’s pooled funds with specific areas of interest. That is because the policies perpetuate the prior generation’s choices as to charitable areas of activity and as to the rate of distribution and accumulation for the future. The advantage that DAF sponsors have over the trustees of an endowed charitable trust is that DAF sponsors are not bound by donor advisor preferences. That is confirmed by the variance power that is reiterated in policies. However, the power is typically expressed as a reserve power, without any procedure or process set out for how the DAF sponsor might be prompted to consider and then apply that power. To the extent that succession options enable donor advisor preferences to be maintained, it would better accord with intergenerational justice if policies set out a procedure requiring the DAF sponsor to consider whether it should use the variance power to amend those preferences after a period that accords a modest degree of respect for the donor advisor. A time period reflecting a generation, say 20–30 years, might be appropriate (see, generally, Murray 2021, 251–254; Lechterman 2023, 110–111).

8 Conclusions

Our review of DAF sponsor policies indicates that they already reflect principles of intergenerational justice in various ways. That is, they do permit but also restrict the accumulation of funds to enable saving for future generations, whilst to some extent also compelling distributions to the current generation (with examples of that compulsion being slightly more emphasized for national DAF sponsors). They also largely maintain a broadly equal degree of decision-making freedom for future generations. However, there is material room for improvement. We have outlined some key areas where improvements could be made such as linking “active” grantmaking to a percentage of DAF account assets, considering minimum distribution requirements at the DAF sponsor level and more formalized periodic consideration of variance powers. On-the-whole, and especially for national DAF sponsors, we expect that such “improvements” would tend to refocus DAF sponsors toward greater distributions to the present generation and to better respond to changed circumstances (such as a depression or COVID-19). Commentators will disagree about the effectiveness of such policies, based in part on factors such as the effective regulation of DAF sponsors and the practical enforceability of policies. However, by sketching the areas in which policies already apply and can be improved, we have shown that such policies can form part of a framework for encouraging donors and DAF sponsors to consider and implement a temporal dimension to their decision-making.