Tax Elasticities of Top Donors

When people talk about “elasticity,” I always nod as if they are actually speaking English but I always secretly think to myself, “what the hell does that mean?”  Context is important.  In this context, “elasticity” is a measure of how charitable giving changes in response to changes in tax rates or rules.  In his recently posted paper, “Tax Elasticities of Top Donors: Evidence from Family Foundations and Donor-Advised Funds”, Simon Essig Aberg measures how much charitable giving by very rich people changes in response to changes in the Tax Code.  I suppose tax elasticity is a good thing to know whenever we change tax rates or eligibility for charitable contribution deductions.  The paper is full of all sorts of statistical models and multiplex demodulations that I can’t comprehend but the conclusions are interesting and understandable.  Why can’t economist speak American, like lawyers?  Anyway, here are some interesting excerpts from the paper:

Top donors face two sets of tax incentives that distinguish them from other donors. First, top donors can often obtain a larger subsidy for their contributions than other donors. Top donors can contribute appreciated assets to their foundation or DAF and avoid paying capital gains tax. Top donors can also time contributions to their foundation or DAF in years when the subsidy is largest, while using their foundation or DAF to smooth giving over time.  Second, foundations face a 5% minimum payout rate—foundations must disburse at least 5% of their assets to charity each year on average or risk paying a steep excise tax. 

. . . 

How do top donors respond to tax incentives? Understanding top donor behavior is an important input to active policy debates. A recent legislative proposal would reduce the immediate tax benefits for contributions and strengthen minimum charitable disbursement requirements. The tax responsiveness of other donors is well-studied (Feldstein and Clotfelter [1976], Feldstein and Taylor [1976], Bakija and Heim [2011]), though the tax incentives of top donors suggest their behavior may differ. This is especially true because the key function of foundations and DAFs relates to intertemporal behavior, and much of the prior literature was constrained to focus on static responses because of data and methodological limitations. In this paper, I quantify how top donors respond to tax incentives, focusing on top donors who give through a family foundation.

Using the foundation tax filings and the Survey of Consumer Finance (SCF), I document four facts about top donor incentives and behavior. First, top donors face stronger tax incentives for contributions than other donors. The average price of giving is 0.78 for top donors and 0.86 for other donors. One reason for this gap is that top donors often contribute appreciated assets and avoid paying capital gains tax, whereas other donors typically contribute cash. The difference in contribution type can explain 31% of the gap in average prices. Differences in marginal tax rates (the median income of top donors is 1.79 times that of other donors) and rates of itemizing deductions (73% of top donors and 59% of other donors itemize) explain the remainder of the gap.

Second, top donors contribute more of securities with greater appreciation. The average noncash contribution [is] two times larger than the average cash contribution. Across years, donors tend to increase their contributions when the average appreciation rate of securities in their portfolio is higher. Within a given year, donors also tend to contribute more of the securities with the highest appreciation rates. These patterns are consistent with the model where donors respond to tax incentives to give.

Third, top donors use foundations to smooth disbursements over time. In the tax filings, disbursements equal zero in 6% of years and contributions equal zero in 78% of years. Anecdotally, many foundations operate by specifying that disbursements equal a fixed proportion of assets averaged over the past few years. 41% of payout rates (disbursements relative to foundation assets) are within 4–6% and 48% of payout rates are within one percentage point of the foundation’s average. In contrast, 65% of contributions relative to foundation assets are within one percentage point of the donor’s average. These results are consistent  with the model where donors bunch contributions to foundations, perhaps to maximize tax benefits, while smoothing disbursements over time.

Fourth, the 5% minimum payout rate affects disbursements. 4% of tax filings report a binding minimum disbursement constraint.  The minimum disbursement may affect behavior even when it is not binding. If a donor’s disbursements exceed 5% of assets one year, then this excess carries over to the next year, lowering the minimum disbursement. Instrumenting for the minimum disbursement with carryover five years prior, I find a strong positive effect on disbursements controlling for fixed donor and year factors. This result is consistent with the model where forward-looking donors exercise precautionary disbursements so the minimum disbursement constraint does not bind in the future.

darryll k. jones