The following article is part three of a Philanthropy Roundtable series examining recent calls for additional regulations on charitable giving and the works underpinning these proposals. To read part one, click here. To read part two, click here.
When it comes to criticisms of donor-advised funds, payout rates are the most common target. Advocates for regulatory changes such as the “Initiative to Accelerate Charitable Giving” argue these personal charitable giving accounts do not payout their charitable assets quickly enough. Just this week, Ray Madoff, a professor at Boston College Law School and proponent of the Initiative, tweeted her frequent assertion that donor-advised funds, or DAFs, “are warehousing huge amounts of funds while nonprofits struggle.” Yet, we know from available data that DAFs overall payout at far higher rates than other charitable vehicles. The divergence between these two points of debate boils down to calculating the ratio of grants to asset value, or the payout rate for DAFs.
Ray Madoff and co-author James Andreoni take on the question of how to best measure DAF payout in a recent working paper, “Calculating DAF Payout and What We Learn When We Do It Correctly.”
While the authors put forward their new formula for calculating payout with great intentions, there are a number of concerns with their proposed approach. As lawmakers consider additional regulations for personal charitable giving accounts, it is important to understand how DAF payout is and should be measured.
All DAF Assets Are Committed to Charity
Before analyzing the payout formulas, it is important to reiterate that every dollar an American puts into a DAF must be directed to a qualified charity and may not be withdrawn for any other use or transferred to a non-charitable account. Over time, all DAFs must have a 100 payout rate. The growing popularity of DAFs is partially attributable to the ability of donors to choose the best timing for their charitable gifts. As we have discussed here before, some wish to pay out contributions immediately, while others wish to wait and allow funds to appreciate, resulting in a larger ultimate gift.
Andreoni and Madoff enter into the measurement discussion with an unclear definition of DAFs, claiming DAFs, “operate in practice like a combination of a charitable investment, savings and checking account.” As the first blog in this series clarifies, DAFs do not operate as a savings account or a checking account. Once funds are put into a DAF, they may only be used for charitable giving.
Measuring DAF Payout
Andreoni and Madoff examine three different calculations for how much of the assets in DAFs are donated to charities each year. With data reported on DAF sponsors IRS Form 990s, the dollar amount paid by DAFs to charities per year is clear. The calculation debate is centered on what share of total DAF assets are represented in this paid-out amount.
The key data available for these calculations are: Beginning of Year (BOY) Assets, Contributions to DAFs, Grants from DAFs, End of Year (EOY) Assets and Returns or Appreciation.
The authors first analyze the formula saying payout equals Grants/End of Year Assets. We agree this formula is misleading, although for different reasons than the authors, which we will get into below.
The second formula considered in the working paper states payout equals Grants/Beginning of Year Assets. This is how the National Philanthropic Trust calculates its data in an effort to match the common methodology used to measure payout rates from private foundations. This allows for an apples-to-apples comparison of payout rates by giving vehicle.
The third formula, which the authors promote, is payout equals Grants/(End of Year Assets + Grants). This formula is based on the questionable assumption that end-of-the-year donations should be included in the assets of a donor-advised fund for the whole preceding year.
A simple example demonstrates why this assumption is flawed. If an individual contributes $200 to a personal charitable giving account on December 31, there’s no way that donor would make enough grants by the end of the year to have the eventual payout accurately reflected in that year’s payout rate. Using the End of Year Assets artificially inflates the denominator and will necessarily skew payout rate lower than it reasonably should be.
The authors correctly say because there are more assets in a DAF at the end of the year than at the beginning of the year, using the formula Grants/Beginning of Year Assets will yield a higher payout rate than using Grants/End of Year Assets. However, this predictable, cyclical timing difference is meaningful, not simply an arbitrary way to calculate how much is leaving DAFs in a year. Donations to DAFs, as to all charities, increase at the end of the calendar year due to the end of the tax year, traditional holiday giving and other preferences. Beginning of Year Assets include the prior year’s contributions and the prior year’s asset appreciation, which donors may use for grant-making.
Yet, even the authors’ favored approach yields a payout rate of nearly 15% – higher than the seven percent minimum proposed in the Initiative.
Why Does This Matter?
The debate over how to calculate payout is ongoing. There are other valuable voices in academia considering the challenge of how to best measure payout. For example, DAF experts H. Daniel Heist at the University of Pennsylvania and Danielle Vance-McMullen of the University of Memphis examine various measures and propose a “flow rate” in their 2019 study.
Regardless of the path this debate takes, as lawmakers face proposals to regulate DAFs more like private foundations, using Beginning of Year Assets in the calculation is necessary to start with a fair comparison to foundation assets. Accepting one working paper as a gold standard without analyzing other measures would be a mistake. One thing is clear: no matter what calculation is used, DAF payout rates fall somewhere between 14.7% and 22.4%. Requiring a seven percent payout rate, as the Initiative calls for, is irrational and would chill charitable giving by sending a signal that is a ceiling, rather than a floor.
To read part one, click here. To read part two, click here.
