Subsidizing Charity: Donations or Endowment Returns?
David M. Schizer1 Draft of May 8, 2016
1 Dean Emeritus and Harvey R. Miller Professor of Law and Economics, Columbia Law School. I appreciate the comments of Louis Kaplow, Kathy Spier, and participants at workshops at Harvard Law School and the Tax Economists Forum. Please do not cite or quote this preliminary draft without permission.
1
Subsidies for charity are supposed to finance positive externalities from charitable activities.2 In principle, the government can pursue this goal by evaluating specific charitable initiatives, calibrating subsidies so they equal the marginal positive externalities from each one.
Although the government sometimes makes this sort of fine-grained judgment,3 this Article focuses on two income tax rules that leave the government very little discretion about which charities to fund: the charitable deduction4 and the exclusion of endowment income.5 Under each rule, as long as charities satisfy very general criteria, federal dollars flow automatically.
Yet notwithstanding this similarity, these rules create very different incentives and effects.
Which of these “nondiscretionary” approaches should we prefer?
This Article identifies important advantages of subsidizing donations over subsidizing endowment income. Compared with endowment income, donations offer more reliable evidence of a charity’s social value. To collect a subsidy for donations, donors have to commit their own money. As a result, the government spends money only if private parties with “skin in the game” commit as well. In contrast, endowment returns are subsidized when charities invest a surplus successfully, but this tells us less about a charity’s value. Admittedly, if surpluses derive from a surge in donations or operating revenue, they reflect favorable judgments of donors and beneficiaries. But surpluses are not always a positive signal. They also can arise, for instance, when charities scale back operations as their missions become less relevant. Even when a surplus is a reliable signal, moreover, investing it successfully is not. After all, investing
requires different expertise than a charity’s mission. In deciding whether the Red Cross is more effective at disaster relief than UNICEF, for instance, we are unlikely to focus on their
investment performance.
Subsidizing endowment returns also creates governance problems, which arise because exempting endowment income encourages donors to front-load contributions. Since charities earn tax-free returns, but donors do not, donors have a tax incentive to contribute assets to charity, instead of investing on their own and donating the return. But transferring assets to the charity creates three governance challenges. First, donors cannot redirect this money to other charities if the cause becomes less relevant. Second, in surrendering control, donors ordinarily impose conditions, which can become dated over time. Third, donors also are less able to monitor agency costs if they give an endowment, instead of keeping assets and periodically
2 A positive externality is a benefit to a third party, which derives from a transaction in which they were not involved. For example, if a graduate of a school gives money to support financial aid, this gift produces positive externalities to students whose tuition is reduced.
3 For instance, various government agencies solicit and evaluate grant proposals.
4 The charitable deduction generally allows donors to avoid paying tax on amounts they give to charity. See Section 170. For example, if a taxpayer earns $1 million of salary and contributes $100,000 to charity, she pays tax on only
$900,000. If her tax rate is 40%, this contribution of $100,000 reduces her tax by $40,000. All references to sections are to the Internal Revenue Code of 1986, as amended.
5 The exclusion of endowment income spares charities from paying tax on income from passive investments. See Section 501(a) (excluding the income of Section 501(c)(3) organizations from tax). For example, a charity that earns $100,000 of interest on bonds it holds does not pay tax on this income, thereby avoiding the 35% corporate tax that otherwise would apply.
2
donating the return. They lose the advantages of staged financing, which requires managers to produce results before receiving the next check.6
Another disadvantage of subsidizing endowment income, instead of donations, is that endowment income is less likely to generate “double” utility. In addition to helping
beneficiaries, charity can offer a second source of welfare gains: a donor’s satisfaction in giving a gift.7 Yet in some cases, donors derive less utility from endowment income than donations. If a donor’s satisfaction comes not just from knowing that beneficiaries are helped, but also from being the one who is helping, donations fill this need. Endowment income sometimes does so as well (e.g., for donors who give an endowment), but does not always. For instance, a donor will not know she is the source of the endowment, and thus will derive no satisfaction from income from this endowment, if she gives a spendable gift that is invested without her knowledge.
In contrast, these double utility and governance problems do not arise in subsidizing donations. Donors and can easily feel a personal connection to each gift. If donors give spendable gifts, they are free to update their assessments of the charity and its mission.
Admittedly, they lose this ability in giving endowment gifts, as noted above. But unlike the exemption for endowment returns, which favors endowment gifts over spendable gifts, the charitable contribution under current law is neutral between these alternatives.8
This is not to say that subsidies for donations are free of problems. Under current law, a familiar issue with the charitable deduction is that it is more generous for wealthy donors. Yet this feature is not inherent in donation subsidies. For instance, the deduction could be replaced with tax credit, which offers the same government match for low-income and high-income donors, while still providing the governance and double utility advantages emphasized above.
This Article’s goal is to analyze these general advantages of donation subsidies, but not to defend every feature of the charitable deduction under current law.
While there are good reasons to subsidize donations, and problems with subsidizing endowment returns, the exemption of endowment income does have an offsetting advantage:
without it, the tax law would encourage charities to spend on current activities, instead of saving for the future. While some commentators actually want the tax law to encourage current
spending over saving, this Article argues that neutrality would be the better posture (at least, if not for the governance problems discussed above). Since charities sometimes have good reasons to save, a tax bias against saving can be counterproductive. To defend this tax bias, some
commentators argue that nonprofit managers have self-interested reasons to save, which the tax law should counter. But in fact, the opposite bias is more likely: a manager’s reputation usually
6 Gilson & Schizer
7 Kaplow.
8 For instance, if the investment returns of both a charity and a donor are taxed, and both endowment and spendable gifts are deductible, a charity ends up with the same money, whether a donor makes an endowment gift or invests herself and contributes the income to charity. See infra Part I.B.2.a.
3
is enhanced more by spending (e.g., on new initiatives, construction projects, etc.) than by setting aside resources for successors to spend. Ultimately, these advantages of neutrality have to be balanced against the governance problems caused by the exemption. In my view, these
governance problems are serious enough to outweigh these advantages. But even so, repealing the exemption is not cost-free.
While other commentators also recommend repealing the exemption, this Article invokes different rationales, which either are not emphasized by others or contradict their analysis. For example, Professor Daniel Halperin also recommends repealing the exemption, but his analysis is different in three ways. First, although governance issues are dispositive for me, Professor Halperin does not focus on them. Instead, the decisive issue for him is whether to favor current or future spending.9 Second, Professor Halperin wants to favor current spending, while my preference is for neutrality. Third, Professor Halperin believes nonprofit managers have a bias against current spending, and I believe they have the opposite preference.
Part I surveys different types of charitable subsidies under current law, and highlights their different effects. Part II emphasizes the advantages of donation subsidies in choosing which causes to support. Part III emphasizes their advantages in monitoring nonprofit managers.
Part IV argues that subsidies for donations are more likely to enhance the satisfaction of donors, and thus to generate “double utility.” Part V considers the best argument for exempting
endowment income: neutrality about whether charities should spend or save, and whether to invest in business assets or passive investments. Part VI shows that donation and endowment subsidies have comparable effects on distribution and deadweight loss, so these factors do not counsel in favor of one or the other.
I. Subsidizing Donations or Endowment Income
At one level, the rationale for subsidizing charities is straightforward: their activities are supposed to produce positive externalities. To keep these social benefits from being
undersupplied, the marginal subsidy should equal the marginal positive externalities from the subsidized activity.
A. Different Ways to Subsidize Charity Under Current Law
Yet obviously there are many ways to channel resources to externality-generating
activities. This Part surveys alternatives under current law, and the rest of this Article focuses on two of them: the charitable deduction and the exemption of endowment returns.
1. Donations
9 Halperin (“ultimately, I believe the appropriate tax treatment of a charity’s investment income should depend upon whether public policy should favor less accumulation and relatively more current spending by charities.”)
4
In 2014, Americans gave $358.38 billion to charity. Donations represented 13.3% of the revenues of public charities.10 Contributing to charity reduce a donor’s taxable income.
Through this deduction, the government partially matches contributions.11
This match is more generous for wealthier taxpayers for four familiar reasons. First, it depends on the taxpayer’s marginal rate.12 Second, taxpayers can claim the deduction only if they itemize, which is something high-income taxpayers are more likely to do. Third, the deduction is especially generous for contributions of appreciated stock, a currency that high- income taxpayers are more likely to use.13 Fourth, the deduction also can be used to avoid estate or gift tax, which are imposed solely on very wealthy taxpayers.14 Congress could change one or more of these features. For example, instead of a deduction, we could use a credit, whose value would not depend on the donor’s tax bracket. Likewise, contributions of appreciated securities could be valued using basis, instead of fair market value. The goal here is not to analyze these features or propose alternatives, since I have covered this ground elsewhere.15 Rather, the goal is to consider why we might want the government to match donations, and how this approach compares with a subsidy for endowment returns.
2. Direct Grants
This Article focuses on government matches through the tax system. Obviously, the government also offers matching grants outside the tax system. Indeed, government fees
represent 24.5% of public charities’ revenue in 2013, and grants add another 8%.16 With most of this money, government officials select which organizations to support. In contrast, this Article focuses on subsidies allocated by private individuals.
3. Tax Exemption for Income
10 The Nonprofit Sector in Brief, http://www.urban.org/sites/default/files/alfresco/publication-pdfs/2000497-The- Nonprofit-Sector-in-Brief-2015-Public-Charities-Giving-and-Volunteering.pdf.
11 See Section 170.
12 For instance, if the marginal rate is 40%, a $1,000 contribution reduces tax by $400. If the marginal rate is 28%, tax is reduced by $280.
13 If a taxpayer purchased stock for $10 per share, and it is trading at $100, the taxpayer can deduct the full $100 fair market value of the stock, but does not have to recognize the $90 of built in gain, which is never taxed.
14 The charitable deduction also has an important limit: taxpayers cannot zero out there full tax liability with charitable deductions. Depending on the type of charity, taxpayers can deduct only 30% or 50% of their adjusted gross income each year, and can carry over unused deductions to other years.
15 Schizer, Subsidizing Charitable Contributions: Incentives, Information and the Private Pursuit of Public Goals, 62 TAX L. REV. 221 (2009).
16 Brice McKeever, The Nonprofit Sector in Brief, http://www.urban.org/sites/default/files/alfresco/publication- pdfs/2000497-The-Nonprofit-Sector-in-Brief-2015-Public-Charities-Giving-and-Volunteering.pdf.
5
In addition to subsidizing donations, the U.S. also supports charities by exempting their income from tax. This exemption and the charitable deduction are not perfectly overlapping.
Some organizations can earn exempt income, even though contributions to them are not deductible. This is the case for social welfare organizations that engage in lobbying, labor unions, and trade associations.17 In principle, then, Congress can subsidize donations to an organization without exempting its income, and vice versa.18
a. Donations
Charities can earn three types of exempt income, and this Article focuses only on the third type. The first are donations. While a donor can deduct contributions, in principle the charity could be taxed on them. As Daniel Halperin has observed, the exemption for donations is no different from the exclusion of other gifts.19 Since this treatment is consistent with the
general rule, it is not a subsidy.20
b. Income from Charitable Operations
Second, any revenue earned by charities in pursuing their charitable missions also is exempt from tax. For instance, no tax is imposed when nonprofit hospitals earn fees from patients and universities receive tuition from students. In 2013, operating revenue represented 47.5% of U.S. public charities’ receipts.21
At first blush, this seems like a subsidy, since a for-profit business’s earnings obviously are taxable. Yet even without the exemption, nonprofits ordinarily would not pay tax on operating revenue, since it funds expenses that usually are deductible. For example, if a
university uses tuition revenue to pay faculty salaries, this revenue is not taxed even at for-profit universities.22
Therefore, exempting operating revenue is a subsidy only if it funds nondeductible expenditures. For example, if a for-profit university uses tuition to buy a building, this expense is capitalized, instead of deducted. As Dan Halperin has observed, the exemption subsidizes
17 Correspondingly, deductible contributions can sometimes be routed to non-exempt institutions – such as foundation grants to for-profit firms – as long as strict limits are imposed on how the money is used.
18 Halperin, Tax L. Rev. 284.
19 If a mother gives $1,000 to her daughter, the daughter pays no tax upon receiving this gift.
20 Halperin, Tax L. Rev. 285; Branch Ministries v. Rossotti, 211 F. 3d. 137, 143 (D.C. Cir. 2000) (IRS represented that Church whose tax exemption was revoked for engaging in political activities could still receive gifts tax-free).
21 Brice McKeever, The Nonprofit Sector in Brief, http://www.urban.org/sites/default/files/alfresco/publication- pdfs/2000497-The-Nonprofit-Sector-in-Brief-2015-Public-Charities-Giving-and-Volunteering.pdf
22 Halperin, Tax L. Rev. 285 (“if an expenditure would be deductible when made, tax exemption for amounts set aside for such expenditures does not reduce the present value of tax payments even if these expenditures are deferred”).
6
capital investments.23 This reality is consistent with Henry Hansmann’s argument for exempting a nonprofit’s income: since nonprofits are not permitted to make distributions to owners, they cannot raise equity capital. Professor Hansmann considers the exemption a substitute for equity capital. The exemption is especially valuable, then, when the nonprofits make capital
investments, which a for-profit would fund with equity capital.
c. Income from Endowment and Other Investments
This Article focuses on a third source of exempt income under current law: endowment income. Instead of using resources for current operations, a charity can invest them. If the money is invested in an active business or is financed with debt, the return can be taxable (as
“unrelated business taxable income”).24 But if the money is invested passively – for instance, in bonds or publicly-traded stock – this investment income also is exempt from tax. In 2013, 4.8%
of income of public charities came from investments.25
Although exempting investment income is the norm in a consumption tax, it is a subsidy in an income tax. In effect, this rule applies a consumption tax to charities, while keeping the income tax as the general rule. When donors invest, their return is taxable, but if they donate these assets, charities earn a tax-free return.
Public charities can accumulate investment assets in three ways. First, donors can require a charity to invest their gift, directing the charity to spend income (but not principal) on a
specified purpose. Second, donors can require their gift to be invested, while giving discretion about how to spend the income. Third, a charity can generate revenue that exceeds its costs, and can choose to invest the surplus. While the term “endowment” technically refers only to the first of these (i.e., restricted spending), this Article uses the term “endowment” to describe all three (i.e., a portfolio of passive investments).
Although the charitable sector had $5.4 trillion of net assets in 2013,26 most charities have little or no endowment. For example, only 1% of charities filing reports with the IRS have endowments of $100 million or more,27 but the assets of these charities represent over $4 trillion, which is nearly 75% of the whole sector’s reported assets. At the other end of the spectrum, 90%
of charities have less than 5 million of assets, and 39% have less than $100,000.
23 Halperin, Tax L. Rev. 285. Professor Halperin also shows that setting aside revenue to fund deductible expenses in the future also is not a subsidy, since the future deduction can shelter this income. Halperin, Subsidy, at 293
24 Section 511. For instance, if an art museum sells science books and city souvenirs in its gift shop, this revenue is taxable. Rev. Rul. 73-105.
25 http://www.urban.org/sites/default/files/alfresco/publication-pdfs/2000497-The-Nonprofit-Sector-in-Brief-2015- Public-Charities-Giving-and-Volunteering.pdf.
26 http://nccsweb.urban.org/tablewiz/showreport.php. Some of these assets presumably are not passive (such as buildings that house the nonprofits).
27 Only 6,837 charities reported assets above $100 million, out of 609,544 charities that filed 990s.
7
Assets of Registered Charities: November 201528 Level of
Reported Assets
Number of Organizations
% of Total Organizations
Total Reported Assets of Cohort
% of Assets Less than
$100,000
239,325 39% 7,633,112,982 .1%
$100,000 to
$249,999
89,772 15% 14,713,264,798 .3%
$250,000- 499,999
67,196 11% 24,180,585,219 .4%
$500-$999,999 59,283 10% 42,465,036,332 .8%
$1 - $5 mil 92,938 15% 210,622,149,550 4%
$5 - $10 mil 22,741 4% 160,586,609,117 3%
$10 - $100 mil 31,574 5% 923,792,940,283 17%
More than $100 mil
6,715 1% 4,022,294,557,549 74.4%
Total 609,544 5,406,288,255,830
Some charities cannot earn investment income tax free, even though they can receive tax deductible contributions. Private foundations, which are nonprofits that do not have enough donors to qualify as a public charity, pay a modest tax on investment income (of 1% or 2%).29
3. Other Subsidies for Charity
Charities also benefit from other subsidies, including exemptions from sales,
unemployment, and property taxes, as well as reduced postal rates. In some cases, they can finance capital improvements with tax exempt bonds.30 These subsidies are beyond this Article’s scope.
B. Different Behavioral Effects
In a world of perfect information and no transaction costs, the mechanics of how subsidies are delivered to charity would not matter. If the relevant activity generates $1,000 of positive externalities, the subsidy should be $1,000.31 How this $1,000 gets to the charity would be unimportant. In these idealized conditions, matching donations should have the same effects as exempting investment returns.
Needless to say, though, these conditions do not hold. Indeed, if they did, a subsidy presumably would be unnecessary, since externalities would already be reflected in prices. Yet the same information and incentive problems that justify a subsidy also lend significance to how
28 Source: National Center for Charitable Statistics, http://nccsweb.urban.org/tablewiz/showreport.php
29 A punitive tax can apply, though, if the foundation violates various legal requirements..
30 Section 145 (allowing exempt organizations to borrow for capital improvements with tax exempt bonds).
31 At the margin, subsidies should equal the marginal positive externalities generated by the subsidized activity.
8
it is delivered. Subsidies for donations create different incentives and behavioral effects than subsidizes for endowment returns. This Section highlights three differences.
1. Encouraging Charities to Attract Donations or to Invest Surplus
First, at a basic level, a deduction for charitable contributions rewards different behavior than an exemption for endowment returns. In essence, contributions trigger one subsidy, while investment returns trigger the other.
Obviously, the charitable deduction is supposed to encourage donors to contribute more to charity. To do so, it lowers the after-tax cost of donations. To benefit from a donation subsidy, then, charities have to persuade people to give money. Although the charitable deduction obviously is claimed by donors (as a reduction in their taxes), its economic benefit goes to charities when it persuades donors to make larger gifts. The deduction does not help
charities that receive no gifts, and are funded solely with income from operations or endowment.
In contrast, the exemption for endowment returns has very different preconditions. It is useful only to charities that generate a surplus, and thus have money to invest.32 This surplus could derive from revenue from operations, government grants, rigorous cost controls, donations of spendable money or, of course, donations to the charity’s endowment. However, if there is no surplus – so the charity spends everything it takes in – the exemption is irrelevant.33
Even among charities that run a surplus, moreover, this subsidy is especially helpful for those who invest successfully. A charity that earns a modest return (e.g., in a bank account) benefits less than a charity that earns more (e.g., in a hedge fund). After all, a tax-free return is more advantageous when this return is high.
The bottom line, then, is that the deduction and exemption reward different behavior.
When one of these subsidies is available, a further question is which stakeholders benefit. For instance, if a university earns a tax-free return, this money can benefit employees (with pay increases), students (with financial aid), or donors (with reduced donations). The incidence of each subsidy – that is, who benefits economically from it – varies with the context. But the threshold question of whether a subsidy is available to begin with depends on which behavior is rewarded. This condition varies for the charitable deduction and exemption.
2. Favoring Endowment Gifts Over Spendable Gifts
A second difference between these subsidies is about the timing of giving. The charitable deduction is neutral between gifts of endowment and spendable money. In contrast, the
exemption for endowment returns encourages donors to accelerate contributions. By applying
32 In theory, the charity can generate resources to invest by borrowing, but debt-financed investments are generally not able to earn tax-free returns; instead their returns are taxed as unrelated business taxable income.
33 Halperin.
9
consumption tax treatment to a charity’s investments, and less-favorable income-tax treatment to a donor’s investments, the rule encourages donors to shift assets to charities, so these assets can generate a tax-free return.
To see this effect, assume Donor is considering either (a) contributing an endowment to Charity; or (b) investing the money herself and contributing the return to Charity each year.
Donor has a $600,000 salary, and Donor and Charity each earn a 7% pretax return on
investments. Is one option more tax-efficient than the other? If endowment returns are exempt, contributing the endowment is more tax efficient. Notably, the charitable deduction does not have this effect; it does not favor endowment gifts over spendable gifts.
a. Parity: Endowments and Donor’s Investments are Both Taxed
First, assume that both donors and endowments are taxed at 40%, and contributions are deductible. On one hand, if Donor contributes a $100,000 endowment, the charitable deduction reduces her after-tax cost to $60,000 (since she avoids $40,000 of tax). Charity earns a7% pretax return of $7,000 per year. Since this endowment return is assumed not to be exempt, Charity pays a 40% tax of $2800, and has $4,200 to spend each year.
On the other hand, Donor can invest the money herself. Since she is no longer
contributing $100,000 to charity, she has to pay $40,000 in tax, and can invest only $60,000. At 7%, this investment generates $4,200 per year. Since Donor contributes this $4,200 to Charity, the charitable deduction shelters it from tax. Therefore, Charity can spend the same $4,200 each year.
In the first case, tax reduces the return on the investment, while in the second case, tax reduces the amount invested. Either way, the result is the same. Therefore, a system that offers only a charitable deduction – with no exemption for endowment returns – does not favor
endowment gifts over spendable support.
b. Parity: Endowments and Donor’s Investments Both Exempt
For the tax law to be neutral between spendable and endowment gifts, the key is to tax donors and charities the same way on investments. This condition is satisfied not only by taxing both, as illustrated above, but also by exempting both.
To see the point, assume the income tax is replaced with a cash flow consumption tax, which allows both the donor and the charity to earn a tax-free return.34 In the above example, if Donor contributes a $100,000 endowment, the charitable deduction allows her to contribute pretax dollars – that is, the full $100,000, without having to pay a $40,000 tax. Charity earns
34 A cash flow consumption tax functions like an unlimited deductible IRA. Taxpayers deduct amounts they invest, and are taxed when they spend this money.
10
a7% return of $7,000 per year, but this return is now assumed to be exempt, so Charity can spend
$7,000 each year.
Alternatively, instead of making an endowment gift of $100,000, Donor can invest the money herself. Yet unlike in the example above, donor can now invest pretax dollars. Under a cash flow consumption tax, she can deduct the $100,000 she invests. This means she can invest the full $100,000, instead of only $60,000. Like the charity, then, donor can generate a 7%
return of $7,000 each year. Under a cash flow consumption tax, this amount would ordinarily be taxed, if it is not reinvested. But Donor can shelter this $7,000 from tax by contributing it to charity. As a result, Charity has the same $7,000 to spend each year.35 The bottom line, then, is that endowment gifts also lose their advantage over spendable gifts if investment returns are exempt for both the charity and the donor.
c. Favoring Endowment: Only Donor’s Investment is Taxed
Yet this neutrality is eliminated if endowments are tax-exempt, but a donor’s investments are taxed. Since the charity has an edge in investing the money, donors can support charities more cost-effectively by donating an endowment, instead of investing money themselves and donating the return.36 Put another way, by applying consumption tax norms only to charities, while using income tax norms for donors, the exemption favors endowments over spendable gifts.
In our recurring example, if Donor contributes an endowment of $100,000 (which, again, is assumed to be deductible), Charity can earn 7% per year after tax, and thus has $7,000 to spend. In contrast, if Donor keeps this $100,000, she can invest only $60,000 after paying a 40%
tax. Assuming this investment yields the same 7% – but on a base of $60,000, instead of
$100,000 – Donor earns $4,200 each year. By making a tax-deductible contribution of this
$4,200 to Charity, she avoids tax on these investment earnings. Even so, Charity has only
$4,200 per year to spend, instead of $7,000.
Although Charity has less money in the second scenario, Donor pays the same price, forgoing $60,000 of after-tax consumption in the first year. Either way, Donor’s cash flows are
35 Notably, this treatment can be replicated – or, at least, approximated – under current law, as long as Donor’s investment can earn a tax-free return. For instance, Donor can fund charitable gifts with an IRA or with appreciated stock. Likewise, if Donor contributes to a private foundation, these assets earn an (almost) tax-free return, since only a modest tax (e.g., of 1% or 2%) is usually imposed on the foundation’s investment income. While Donor has given away this money, she still can control how it is invested, and she retains the right to choose which operating charities the foundation supports. In other words, a foundation functions as Donor’s personal charitable endowment.
36 Halperin (“because other future consumption by the donor is affected by the income tax on investment earnings accumulated for this purpose, when the charity is exempt, the price reduction for charitable spending, as compared to other deferred consumption, is relatively greater the longer consumption is deferred”).
11
the same: $300,000 for consumption in the first year, and $360,000 thereafter.37 The main difference, obviously, is that contributing up-front allows the full $100,000 to be invested, while keeping this money means only $60,000 can be invested.38
Notably, endowment gifts retain this advantage over spendable gifts even if the charitable deduction is repealed. Returning to our example, if Donor wants to use $100,000 of pretax income to contribute an endowment, she can contribute only $60,000 if the contribution is no longer deductible. Charity earns a 7% return of $4,200 each year, and can keep it all (since endowment returns are assumed to be exempt). In contrast, if Donor invests this $60,000 herself, she earns a pretax return of $4,200. But if charitable contributions are not deductible, she has to pay a 40% tax on this income, and has only $2,520 to contribute. So if Donor keeps the
investment, Charity has 40% less to spend each year ($2,520, instead of $4200). In other words, the exemption favors endowment gifts over spendable gifts, regardless of whether contributions are deductible.
3. Current or Future Spending by Charities
The exemption for endowment returns affects not only the time preferences of donors – favoring endowment gifts over spendable gifts, as noted above – but it also the time preferences of charities. For a charity – as for any taxpayer – a tax on investment returns can influence the motivation to save. There is both an income and a substitution effect, which cut in opposite
37 The following tables compare the cash flows of Donor and Charity in both scenarios:
Endowment Gift Year Salary Investment
Income
Donation to Charity
40%
Tax
After Tax Income
Personal Invest- ment
Personal Spending
Charitable Spending
1 600k 0 100k 200k 300k 0 300k 0
2 600k 0 0 240k 360k 0 360k 7k
3 600k 0 0 240k 360k 0 360k 7k
4 600k 0 0 240k 360k 0 360k 7k
5 600k 0 0 240k 360k 0 360k 7k
Spendable Gift Year Salary Investment
Income
Donation to Charity
40%
Tax
After Tax Income
Personal Invest- ment
Personal Spending
Charitable Spending
1 600k 0 0 240k 360k 60k 300k 0
2 600k 4200 4200 240k 360k 0 360k 4200
3 600k 4200 4200 240k 360k 0 360k 4200
4 600k 4200 4200 240k 360k 0 360k 4200
5 600k 4200 4200 240k 360k 0 360k 4200
38 Either way, the investment return can be earned tax-free. The endowment return is tax-free because of the tax exemption, while the private investment return is tax-free because the deduction shelters the income from tax.
12
directions. On one hand, if charities have to accumulate a particular amount – for instance, to buy a building – they have to save more to reach this target. On the other hand, since a tax on endowment returns increases the cost of saving – but not the cost of spending now – it favors current activities over future ones.
Relatedly, if earnings from a charity’s operations are exempt, but earnings from its passive investments are taxed, charities are encouraged to favor (exempt) business assets over (taxable) passive assets. For instance, a charity that receives a $10 million donation would be better off buying a building to house its operations, instead of an endowment to fund future expenses.
Notably, the charitable deduction (on its own) does not favor one of these choices over the other. After all, donors receive the same deduction, whether their gift finances a building or a portfolio.
Exempting endowment returns from tax eliminates these distortions. Again, the
exemption applies consumption tax norms when charities save and invest. As Dan Halperin has observed, “an exemption for investment income may be necessary to make the charity neutral both as between current and future spending and as between direct ownership of a business and a diversified portfolio of stocks.”
***
To sum up, although the charitable deduction and the exemption for endowment returns both channel government money to charities, this Part has highlighted three key differences.
First, the deduction rewards contributions, while the exemption rewards investment returns.
Second, the deduction is neutral about whether donors give endowments or spendable gifts, while the exemption favors endowments. Third, without the exemption, charities would have a tax incentive to favor current spending over future spending, as well as assets for their operations over passive investments. In light of these different effects, which subsidy is better? The rest of this Article considers strengths and weaknesses of subsidizing donations, on one hand, or
endowment returns, on the other.
II. Picking the Right Cause: A Governance Advantage of Subsidizing Donations
Which type of subsidy is more likely to flow to charities that add the most social value?
In a world of perfect information and incentives, we could rely on the government to make this choice. For most of the money it spends, the government is supposed to make this sort of judgment, but charitable subsidies are different. The government does not choose which charities to support, so we need another way to channel resources to high-value charities.
Subsidies for donations are more likely to do so than subsidies for endowment returns, since the government can piggyback on the judgments of donors who have “skin in the game.”
13
A. Government Officials Do Not Allocate Charitable Subsidies
As long as charities satisfy very general requirements, subsidies for donations and
endowment returns are automatically available. For instance, charities have to be “organized and operated exclusively for religious, charitable, scientific, testing for public safety, literary, or educational purposes.”39 If this general condition is satisfied (along with some others), public money flows, as long as donors are willing to fund the charity (in the case of the charitable donation) or the charity’s endowment earns a positive return (in the case of the exemption).
Likewise, government officials do not have discretion to adjust the size of the subsidy.
Regardless of what government officials think of the cause, the subsidy rate is based on marginal tax rates: the charitable deduction uses the donor’s rate, while the exemption for endowment returns uses the corporate rate, which would otherwise govern a nonprofit’s investments.
Obviously, this is a blunt way to scale the subsidy, which in theory should vary with the marginal externalities of the subsidized activity.
1. Advantages of Not Relying on Government Officials to Allocate Subsidy
Even so, this lack of substantive government oversight is a feature, not a bug. It has four advantages. First, the lack of government control allows charities to be more flexible and nimble. They can be launched quickly. If conditions change, charities do not need government approval to adjust priorities and strategies.
Second, for the same reason, they are more independent politically. Even if there is no political consensus for a nonprofit’s mission, it still is able to act, and to tap government funds.40 Unlike legislatures, nonprofits can advance causes that are “cutting edge” and have not (yet) attracted the median voter’s support. It is no accident that the civil rights, women’s rights, and environmental movements all began in nonprofits. The government did not embrace them until popular support became more widespread.41
Third, the independence of nonprofits also allows them to fill gaps when direct government involvement would compromise important values. For instance, the government should not choose which religions to support. The same is true of newspapers, since they are supposed to monitor (and criticize) the government. For these institutions, charitable subsidies offer government funding without government control.42
Fourth, since charities do not depend on the good will of government officials, they are freer to compete with the government – and each other – to develop the best solutions. In promoting competition, nonprofits function like for-profit firms, enhancing quality, imposing
39 Section 501(c)(3).
40 Weisbrod
41 Schizer, Subsidizing Charitable Contributions
42 Schizer, Subsidizing the Press
14
discipline, and promoting innovation. Admittedly, government can simulate this competition on their own, at least to an extent, by tasking different agencies or levels of government to act independently. For this reason, Brian Galle argues that charitable subsidies are more useful for missions that state and local governments cannot undertake.43 Yet even when there is some competition already, the charitable sector can add more.
2. Disadvantages of Not Relying on Government Officials to Allocate Subsidy Although the lack of substantive government oversight offers these advantages, it has three familiar downsides as well. First, the government provides no coordination or quality control. When separate charities pursue overlapping missions, they might duplicate effort and forgo economies of scale. Likewise, each donor decides what to support without full information about what others are supporting.44 In addition, donors are free to support – and direct public money to – misguided causes and poorly run organizations.
Second, some commentators question the legitimacy of relying on private individuals to allocate public money, instead of democratically elected representatives.45 Yet although Congress doesn’t allocate these funds itself, it has authorized the way they are allocated by enacting the relevant subsidies. Delegating the allocation of funds is not an unusual step for Congress. Agencies have some discretion over their budgets, panels of experts allocate some grants, and the like. A related process-based concern is that wealthy people exert
disproportionate control over charitable subsidies, since they have the capacity to make larger contributions.46 Yet this problem is not unique to charity. Through campaign contributions, wealthy people also have added influence with elected officials.
Third, the influence of wealthy donors raises concerns about outcomes, as well as process. If these donors have idiosyncratic preferences, they may route public money to causes that lack mainstream appeal. For instance, wealthy donors give less to religious organizations than other Americans, and more to cultural and educational institutions.47 Even so, wealthy people are not monolithic in supporting particular causes. George Soros funds causes some that the Koch brothers oppose, and vice versa. To a significant extent, wealthy people have values and preferences as diverse as the public at large.
B. Subsidizing Donations: A More Reliable Delegation
43 Likewise, Eric Posner and Anup Malani have argued that for-profit corporations should also be tasked with pursuing public goals, and arguably should be subsidized when they do so, since they have the advantages of equity capital and the discipline associated with the profit motive. In fact, a charitable subsidy for for-profit firms is more feasible than Posner and Malani seem to suggest. For instance, foundations can make grants to for-profit firms as long as the grant has “expenditure responsibility” language requiring the for-profit to use the funding solely for a specified charitable purpose and to account for the way it uses the funds.
44 Levmore
45 Ilan Ben Shalom.
46 Ilan Ben Shalom
47 Andreoni, Philanthropy (2004) 50-51.
15
Since commentators assign different weights to these competing advantages and disadvantages, their enthusiasm for these nondiscretionary subsidies varies. Likewise, commentators also disagree about the effectiveness of the main alternative – relying on
government officials to allocate money – and thus about interest group influence, as well as the government’s limited information and expertise.
Even so, skeptics and enthusiasts should agree that if a nondiscretionary approach is used – that is, if charitable subsidies are dispensed with general criteria – these criteria should
correlate, as much as possible, with a cause’s social value. In other words, more valuable initiatives should receive more funding.
On this dimension, subsidies for donations are more reliable than subsidies for endowment returns for two reasons. First, attracting contributions is a better proxy for a charity’s social value than running a surplus and investing it successfully. Second, spendable gifts are more flexible than endowment gifts, and thus allow charities to adapt more easily to changing conditions.
1. Contributions as a Signal of Social Value
With the charitable deduction, government money flows only if donors value a charity enough to support it themselves. In fact, to qualify as a “public charity” instead of a private foundation – and thus to enjoy a range of other benefits – charities have to draw support from a sufficiently broad pool of donors.
Obviously, this is a competitive process. Since giving charity means forgoing consumption, donors have to value spending on the cause more than marginal dollars for themselves. Donors also have to believe their cause is more worthwhile than others. After all, the range of options is vast, including religious organizations, poverty and disaster relief, education, public policy, civil and human rights, the environment, cultural organizations, and much more. To be competitive, a charity has to make a strong case.
This is all the more true because donors have access to significant information.
Organizations have websites and printed materials, and also sponsor events to publicize their activities. For sufficiently large gifts, a charity’s managers spend time with potential donors. In many cases, donors give time as well as money, and their volunteer work educates them about the charity’s strengths and weaknesses. The media also covers nonprofits, and various websites evaluate their work.
This is not to say that every charitable decision is based on rigorous analysis, rich
information, and deep insight. Coworkers, neighbors, and family members ask each other to buy raffle tickets or cookies for their causes. Some donors buy tickets to a charity’s dinner because a friend is being honored, not because they personally value the cause. Some contributions are given to satisfy a social or professional obligation, or to avoid the awkwardness of saying “no.”
16
But even casual gifts can be a positive signal in some cases. Donors may support their friends’ causes with the expectation that friends will reciprocate; if so, these gifts really are indirect support for the donors’ own causes. Donors may support a family member or friend’s cause because they respect her judgment. In any event, these uninformed choices are more likely for small gifts. For larger gifts, donors are likely to invest more time and thought.
Even when donors take these decisions seriously, their choices will not appeal to
everyone. Donors have heterogeneous preferences and values, so causes that resonate with some are unpersuasive – or even unappealing – to others. But charitable subsidies are supposed to support causes outside the mainstream, as noted above.
Admittedly, some donations – and, thus, some uses of government money – are misguided. This can happen not only because donors are uninformed or have different preferences, but also because they sometimes have poor judgment. In some cases, nonprofit managers may serve as a constraint. If they consider the donor’s idea unwise, they will try to persuade the donor to support something else. If this effort fails, the manager may refuse the gift, for instance, if the charity would have to devote matching money, or if the manager worries that the initiative would undermine the charity’s (and her personal) reputation. Yet even if the nonprofit takes the gift – for instance, because the nonprofit manager also has flawed judgment – this problem is hardly unique to charity. Unwise choices also occur in the for-profit sector, skewing resource allocation and shrinking the tax base, since tax losses are deductible.
Likewise, unwise votes and campaign contributions distort government policy. It is not obvious why this problem is worse in charities than anywhere else.
The bottom line, then, is that a charity’s ability to raise money ordinarily should correlate with its social value. Although this correlation is not perfect, it should be meaningful. Since donors have “skin in the game,” the government can be more comfortable piggybacking on their judgments.
2. Endowment Returns: A Less Reliable Signal of Social Value
In contrast, charities have to satisfy very different preconditions to claim a subsidy for endowment returns. First, the charity needs to generate a surplus, so it has money to invest.48 Second, the investment needs to be successful, since sheltering a return from tax obviously becomes more valuable as the return increases.
Although an investible surplus can be a signal of quality, this is not always the case. On one hand, a surplus that derives from increasing revenue – whether from current donations or from fees for the charity’s services – shows that donors and beneficiaries value the charity’s work.
48 In theory, the charity can generate resources to invest by borrowing, but debt-financed investments are generally not able to earn tax-free returns; instead their returns are taxed as unrelated business taxable income.
17
On the other hand, an endowment donated long ago reveals more about the charity’s value in the past than today. The donors who gave this money are not offering a current judgment. In some cases, these donors – and the managers who secured their support – passed away years ago. While endowments are permanent, a charity’s value can change. The cause may be less relevant, and the organization may not run as well. Yet once donors have given an endowment, they can no longer reevaluate the charity and redirect their support.
This concern highlights a potential advantage of endowment gifts to private foundations, instead of operating charities: in allocating the foundation’s income every year, donors can make updated judgments about which charities to support.
In any event, another reason why an operating charity’s investible surplus is an imperfect signal of quality is that it can arise not just from increasing revenue, but also from decreasing expenditures. While cost-cutting can be a positive signal – for instance, as a sign of discipline and efficiency – the opposite can also be true. Perhaps the mission has become less relevant, so there are fewer opportunities to spend money effectively. Or perhaps managers are conserving cash to preserve their jobs, since donors have stopped funding the cause.
Even if a surplus is evidence of social value, investing this surplus successfully is not.
After all, why would a charity’s ability to choose the right stocks and bonds correlate with the value of its mission? Earning robust investment returns is not as dependable a proxy for social value as attracting donations.
Admittedly, a concern about donors allocating public money, discussed above, is that wealthy donors wield disproportionate influence. Yet this concern is also present with
endowments, especially those that derive from endowment gifts, instead of operating surpluses.
After all, endowment gifts are likely to come from wealthy donors, since they have enough assets to front-load their giving.
3. Flexibility
Compared with endowment gifts, spendable gifts not only entail a current judgment, but also offer charities more flexibility. Indeed, they avoid two risks posed by endowments: first, endowments can undercut a charity’s motivation to change with the times; second, endowments can also undercut its ability to do so.
After all, charities are forced to remain relevant if they depend on operating income and spendable donations. Otherwise, they can lose their funding. These charities depend on “staged financing,” much as startups funded by venture capitalists do.49 Since these charities are
constantly being evaluated, they have to keep demonstrating the value of their mission. In contrast, charities that live off endowments are not under the same pressure. With less need for
49 Gilson & Schizer
18
new donations and operating income, these nonprofits might become complacent, addressing dated problems with stale solutions.
In addition to undercutting a charity’s motivation to change, endowments can undercut its ability to do so by restricting how it can spend money. Restrictions that were sensible years ago may no longer make sense. The cause may be less relevant, or the right strategies for advancing it may have changed. Admittedly, spendable gifts can also be restricted, but these limits are imposed currently. Therefore, donors can tailor the restrictions to current circumstances, and also can modify these limits. Yet these steps are less feasible for endowments – and sometimes are impossible – especially when the relevant donors have passed away.
III. Monitoring Management: A Governance Advantage of Subsidizing Donations Charitable subsidies should flow not only to the right causes, but also to the right
organizations, favoring those that run efficiently. Yet just as government officials do not choose which causes to fund, they also do not evaluate nonprofit managers. A different mechanism is needed to discipline nonprofit managers and reduce agency costs. This brings us back to the advantage of subsidizing donations, instead of endowment returns: for this issue as well, the government can piggyback on the judgments of donors.
A. The Challenge of Monitoring Nonprofit Managers
A familiar challenge in advancing public goals – in both nonprofits and in government – is the difficulty of measuring success. When the goal is to promote positive externalities, profitability obviously is not a reliable proxy. This means success can be especially difficult to measure in nonprofits and government.50
Therefore, agency costs can loom especially large. When managers’ performance is hard to assess, they become freer to define their mission or workload in self-interested ways.
Incompetence also becomes harder to detect. These challenges are all the more daunting because nonprofits do not have owners to monitor managers.
Indeed, a key risk is that extra resources for charities – whether from donations,
endowment income, or government subsidies – will not actually help beneficiaries. For instance, contributions to hospitals or soup kitchens might fund above-market pay to managers, instead of enhancing the services offered to beneficiaries.
50 If beneficiaries cannot effectively measure quality, they may worry the organization is taking advantage of them.
Henry Hansmann has argued that nonprofit form – and, in particular, the inability to distribute profits – can build trust by eliminating a key incentive to skimp on quality. Yet the nondistribution constraint is only a partial solution.
Even though managers cannot distribute profits, they can access the surplus in other self-interested ways. For instance, they can claim excessive pay, exert insufficient effort, or pursue pet projects.
19
Even so, agency costs at nonprofits are subject to three constraints, whose effectiveness varies with the context. The first is product market competition. For example, universities compete for students and hospitals compete for patients. A second limit is the internal motivation of managers. Ordinarily, they work at nonprofits because of a commitment to the cause.51 To an extent, then, nonprofit managers should be less interested in taking personal advantage of their position. Yet even idealists sometimes add less value (and are less competent) than they realize. They also may define their responsibilities in ways that are better for them personally than for the organization. Therefore, a third constraint on agency costs at nonprofits – the competition for donors – can be quite important. In deciding whether to support a charity, donors assess whether it is well run.
Realistically, small donors are unlikely to impose much discipline. Given their modest stake, they may not be motivated to dig into the details of a charity’s operations. Like small shareholders in for-profit firms, they also have only limited influence on managers.
Yet large donors are different. When negotiating a significant gift, they have the incentive to probe details of a charity’s operations, and also have significant influence over management. In addition to constraining agency costs, donors also can add value by sharing expertise, for instance, about running a complex organization.52 Given the value of this input, a subsidy that encourages it has advantages over one that does not.
Even so, not all forms of donor influence are constructive. For instance, donors
sometimes seek to help themselves, instead of the cause. In these cases, donors are a source of agency costs, instead of a constraint on them. In response, charitable subsidies have familiar rules to police donor self-interest, such as the “private benefit” doctrine.53 Although not a complete solution, these limits are helpful.
To sum up, then, monitoring quality and policing managerial agency costs are as
important with charities as with for-profit firms – indeed, perhaps more so. Therefore, subsidies that promote socially valuable monitoring have advantages over subsidies that do not.
B. Donation Subsidies Encourage Donors to Monitor Nonprofit Managers
The challenge considered in this Part is somewhat different from the challenge in the last Part – subsidizing the right management, instead of the right cause. Yet the subsidies’ relative merits are the same: subsidizing donations allows the government to piggyback on donor judgments, but subsidizing endowment returns does not.
1. Donations and Endowment Returns as Evidence of Low Agency Costs
51 Daniel N. Shaviro, Assessing the “Contract Failure” Explanation for Nonprofit Organizations and Their Tax- Exempt Status, 41 N. Y. L. Sch. L. Rev. 1001 1996-1997.
52 Schizer
53 This dynamic is like the potential and challenge associated with large shareholders of private firms, who can discipline management but also may seek private benefits of control. See Goshen & Hamdani.
20
With the charitable deduction, government money flows only if donors are willing to support a charity. In making this choice, donors can evaluate not only the cause, as emphasized above, but also the management. If donors consider management inept or self-interested, they can withhold support. To “stay in business,” then, nonprofit managers have to produce results that impress their funders. Therefore, staged financing can enhance the performance of nonprofits, just as it disciplines high tech startups.
Subsidies for endowment returns, by contrast, are not conditioned on donor judgments.
To collect this subsidy, charities have to satisfy two conditions, as noted above: running a surplus and investing it successfully.
Admittedly, these conditions are sometimes a reasonable proxy for managerial
competence. For instance, a surplus may derive from robust operating revenue or gifts, and thus from favorable assessments of beneficiaries and donors. When the surplus stems from gifts, the timing of the gift is important. Current gifts, which entail contemporaneous judgment about managerial competence, are more reliable evidence than past gifts. The ability to invest effectively shows competence of a sort. But this skill does not necessarily correlate with the main responsibilities of a nonprofit’s managers: to advance the charity’s mission.
A surplus also can derive from choices that reflect badly on management. For example, a hospital might overcharge unsophisticated patients, while a soup kitchen might turn away needy people. To make matters worse, managers might be hording cash to protect their jobs, so the payroll is covered even if disillusioned donors end their support. The fact that an organization has robust endowment returns, then, is not a reliable signal about agency costs.
2. Spendable Gifts as a Constraint on Agency Costs
Endowments are not just imperfect evidence that a charity is well run. They also are a potential impediment, since they insulate managers from the need to produce results. Although donors evaluate managers before giving an endowment gift, this due diligence is conducted ex ante.54 Once donors have given an endowment gift, they forgo the right to reevaluate the charity over time. Donors who contribute endowments may still wield influence in other ways, such as by serving on the nonprofit’s board, but they lose the ability to redirect their giving.
In contrast, if donors invest the assets themselves and donate the return each year, they can withhold support if managers underperform. Just as the need for spendable gifts pressures managers to change with the times, as noted above, it also can motivate them to run
organizations more efficiently and to take less for themselves. For example, Brian Galle and David Walker have found that university presidents earn less “at institutions that are more highly dependent on current donations as a source of revenue (versus tuition, grants, etc.).”55 This may
54 Testamentary gifts – and, thus, the charitable deduction in the estate tax – also front-load monitoring in this way.
55 Brian Galle & David Walker, Nonprofit Executive Pay as an Agency Problem: Evidence from U.S. Colleges and Universities, 94 B.U. L. Rev. 1881 (2014).
21
seem counterintuitive, since the president presumably should be rewarded for effective fundraising. But Professors Galle and Walker believe this finding shows that spendable donations constrain agency costs.56
Of course, spendable money does not always enhance the efficiency of nonprofits.
Managers funded by spendable gifts spend more time raising money, leaving less time for other responsibilities. Moreover, when managers are more informed than donors, and thus make better judgments, donor influence can lead to inferior decisions. Donors also can be a source of agency costs in some cases.
Given this tradeoff, setting the right balance can be a challenge, and is likely to vary with the context. But for our purposes, there is no need to resolve this issue. Rather, the question is whether the tax law should favor endowment gifts over spendable gifts, as under current law.
Given the tradeoffs here – and, indeed, the notable advantages of spendable gifts in many contexts – this tilt toward endowment is undesirable.
3. Private Foundations: A Solution and a Problem
Admittedly, there is a way under current law to get most of the tax advantages of
endowments, as well as some of the governance advantages of spendable gifts. Instead of giving endowments to operating charities, donors can give them to private foundations. Although a foundation’s investment returns are subject to a 1% or 2% excise tax, this rate obviously is much lower than the donor’s tax rate. The donor then can decide each year which charities to support with the income. This ability to redirect giving can push charities to change with the times, and can motivate managers to perform.
Yet although foundations can help solve agency costs, they also are a familiar source of agency costs – at least, when they are run by a professional manager, instead of the donor.
While a foundation manager is supposed to monitor charities, who is monitoring her? After all, a foundation manager might define her responsibilities and compensation in self-interested ways.
In addition, she might substitute her own preferences for the donor’s, especially after the donor’s death. It is commonly observed, for instance, that the Ford Foundation’s activities no longer reflect the preferences of its founder. Therefore, private foundations can mitigate some agency costs, while exacerbating others, depending on the context.
IV. “Double Utility” Advantage of Subsidizing Donations
The last two Parts considered whether charitable subsidies are likely to flow to the right organization. The focus was on how charitable dollars are spent. A different issue is how these dollars are raised. As Louis Kaplow has observed, a distinctive feature of charity is that it can
56 Id. (“The fact that more powerful donors are able to drive down pay levels implies that presidents at schools with less-influential donors are extracting more pay than donors would want.”)
22
