Social Research: An International Quarterly

AN OVERVIEW OF CHARITABLE GIVING IN THE UNITED STATES

It is often said that Americans are the world's most generous citizens. The claim is based upon the large amount of money given away every year by Americans, approximately $315 billion in 2012, a figure that, except during recessions, has risen more or less constantly for the past 40 years (see figure 1 for the 40-year trend line).

Click for larger view
View full resolution

Figure 1.

Total giving, 1971–2011 (in billions of dollars).

The United States has many more citizens than other highly industrialized countries, so the absolute amount given is not especially significant. The better data point to substantiate the claim that Americans are especially generous is the total amount donated as a percent of GDP, about 2 percent, a figure that has remained more or less constant over 40 years, even during recessions (for a general overview, see Reich et al. 2011; see figure 2 for a graph of giving as a percent of GDP). [End Page 104]

Click for larger view
View full resolution

Figure 2.

Giving as percent of GDP.

Click for larger view
View full resolution

Figure 3.

2012 Charitable contributions, by source of contributions (in billions of dollars). Total dollars contributed = $316.23 billion. Source: Giving USA 2013.

[End Page 105]

Four primary sources account for the more than $315 billion donated in 2012: living individuals, bequests by individuals at their death, corporate foundations, and philanthropic foundations (see figure 3). Gifts by living individuals account for the vast majority of donations (72 percent), reflecting the fact that the incidence of charitable giving in the United States is indeed widespread. According to recent studies, nearly 90 percent of American households make charitable donations (Havens, O'Herlihy, and Schervish 2006, 543).

But how is this considerable sum of charitable giving distributed? What charities do donors favor? The theme of this volume is "caring for the needs of strangers." How much of charitable giving in the United States goes to provide for the basic needs of others? The perhaps surprising answer: not much at all. For anyone who believes that charity implies something about alms-giving or assistance for the disadvantaged, the sunny picture of American charitable giving here becomes cloudy.

Consider first this simple chart (figure 4) of the distribution of charitable dollars in 2012 by type of recipient organization. What stands out immediately is that the largest class of recipient organization is religion, meaning giving for the sustenance of religious groups (churches, synagogues, mosques), for their facilities, their operating costs, and their staff salaries. In this sense, religious groups look less like public charities and more like mutual benefit societies.

Some observers, like the Princeton sociologist Robert Wuthnow, argue that at least some of this religious giving should be understood as assisting the needy. But Wuthnow, who writes admiringly of faith-based social services, nevertheless finds that "the amount spent on local service activities is a relatively small proportion of total giving, probably on the order of 5 percent" (2004, 49). Moreover, faith-based social service organizations, such as the Salvation Army or Mercy Corps, have been counted in the relevant service provision category. Bottom line: one-third of charitable giving in the United States—roughly $100 billion—is support for religion. [End Page 106]

Click for larger view
View full resolution

Figure 4.

2012 giving, by type of recipient organization. Total = $316.23 billion. Source: Giving USA 2013.

What of the remaining $200 billion, or two-thirds of charitable giving? Giving to education, health, and human services are the next largest recipients. A portion of the donations in these categories provides for basic needs such as scholarships for the poor and charity care, but many donations are to higher education to support the donor's alma mater or to capital campaigns for hospitals. Still other categories, such as giving to the arts, culture, and humanities, do not appear to constitute support for the needy. A simplistic reading of figure 4 leads to the conclusion that a small percentage, surely less than half of all charitable giving, goes to the needy. But let us acknowledge that it is hard to determine what portion of giving in any category constitutes giving to assist the needy. [End Page 107]

Click for larger view
View full resolution

Figure 5.

Estimates of giving focused on the needs of the poor, 2005. Total giving = $252.55 billion.

We have to turn to other studies for a more precise answer. In 2007, Indiana University's Center for Philanthropy used multiple data-sets, using generous assumptions, to determine what amount of charitable giving is directed to the neediest (see "Patterns of Household Charitable Giving by Income Group, 2005" at https://scholarworks.iupui.edu/items/3dd60e3c-f721-4349-ae40-2c56be3f45f9). The authors find that only one-third of charitable giving goes to the poor for the provision of basic needs (see figure 5 for a pie chart, reproduced from data in the 2007 report).

A closer look at the patterns of giving by living individuals, broken down by income, reveals that the inclination to give for the purpose of meeting basic needs declines as one rises up the income ladder. According to estimates, in 2005 millionaires contributed only 4 percent of their giving to basic needs, contrasted with donors with [End Page 108] less than $100,000 of income devoting more than 10 percent of their giving to meeting basic needs. There are eye-popping differences, too, in giving to education and health: millionaires devoted more than 50 percent of their philanthropy while those with less than $100,000 of income gave only 6 percent to these categories. While those with less than $100,000 of income are the most generous in giving to basic needs, it is nevertheless surprising that even here total giving, as a percent of income, amounts to only 10 percent. This group of donors gives overwhelmingly to religion—two-thirds of their total giving—and as we have just seen, very little of religious giving can be construed as giving for the basic needs of others. By contrast, the wealthiest give only 17 percent of their charity to religion.

Overall, the portrait of charitable giving intended as providing for the poor is not flattering. The Indiana study concludes by making very generous assumptions about what portion of giving in other categories, such as giving to education or health, might benefit the poor.¹ In the "best case" scenario, the authors find that at most one-third of all charity goes to meet basic needs.

What can we conclude from these data? The lesson is obvious: if we think of philanthropic or charitable giving as having something to do with caring for needs of strangers, the actual practice of giving in the United States does not meet the test. As an empirical matter, philanthropy is not much about caring for the needs of strangers. To the contrary, the great majority of American charity is support for religious organizations, with support of education and health trailing far behind, and assistance for the needy lagging still further behind.

It is not my purpose here to explore why in the United States charity so infrequently lives up to the traditional connotation of the word "charity": alms-giving and assistance for the disadvantaged. Among the explanations, however, I would include the following: the nonprofit sector in the United States is wildly permissive, oversight of the sector is lax, and approval of nonprofit status is trivially easy to obtain. The United States has more than 1.3 million public charities, not counting religious organizations, which receive but do not [End Page 109] need to file for nonprofit status, and more than 50,000 new public charities have been created every year for the past decade (see Reich, Dom, and Sutton 2009). No other country in the world makes it as simple to create a tax-exempt nonprofit organization. Practically the only requirement for 501(c)(3) status—American tax terminology for status as a public charity—is that revenue not be distributed as profit to shareholders or investors. The mission or purpose of the nonprofit organization is effectively irrelevant. A glance at the definition of a public charity under United States law reveals the expansive array of sanctioned organizational missions: exempt purposes include "charitable, religious, educational, scientific, literary, testing for public safety, fostering national or international amateur sports competition, and preventing cruelty to children or animals" (see IRS Publication 557, available online at https://www.irs.gov/publications/p557/ch03.html#en_US_201l_publink1000200093).

Philanthropy appears to be more about the pursuit of one's own projects, a mechanism for the expression of one's values or preferences rather than a mechanism for redistribution or relief for the poor.

Is this worrisome? From the standpoint of the assumption I take to be embedded in the conference and in this volume—caring for the needs of strangers—giving patterns in the United States cannot be seen as especially praiseworthy. But perhaps caring for the needs of strangers is not the task of charity or philanthropy. Perhaps it is appropriate, or at least permissible, that giving has a broader mission, a more capacious scope, than providing for the poor.

Social scientists have studied the empirical landscape of charity in the United States with some care, but relatively little has been written about fundamental normative questions concerning the purpose of charitable giving in democratic societies. What is the role of philanthropy or charity? What is the relationship between philanthropy and justice? What principles should guide the creation of public policies that govern charitable giving? [End Page 110]

One might argue, for instance, that charity and justice are in tension. Supposing social justice were achieved, would not the need for charity drop precipitously? If one believes that meeting the basic needs of citizens were a matter first and foremost of justice, a responsibility of just social institutions and arrangements, not voluntary charity, then a just state would not have much need for charity. The more charity we have, the greater the sign that conditions of social injustice abound.

Political philosophers occasionally address such questions, but rarely with any reference to the contemporary practice of charity (see for instance Kymlicka 2001 and Walzer 1992). With the portrait of charitable giving in the United States now drawn, my aim is to take up one particular normative question: What justification could there be, if any, for the provision of tax incentives for philanthropic giving? Should tax incentives for charitable giving prioritize giving to assist the needy? Should the incentive reward charitable giving of any and all kinds? Or should we do away with tax incentives for charity?

These are questions, I believe, of both philosophical interest and public policy relevance. In the wake of the great recession of 2008, we have entered an era of government austerity and the possibility of fundamental tax reform in the United States. The Obama administration has proposed nearly every year to cap the charitable contributions deduction for the wealthiest at 28 percent. The Romney campaign proposed to cap all deductions at a set limit, a proposal that would have gutted the charitable giving incentive. Bipartisan budget plans, such as the Simpson-Bowles proposal, have also suggested radical changes to the treatment of charitable giving. It is far more pressing in 2013 to articulate the justification for charitable subsidies than in the past.

Economists have answered the call. Jon Bakija's essay in this volume examines tax incentives for charitable donations, assessing whether such incentives succeed in stimulating donations that would not be made in the absence of the incentives (Bakija 2013). Bakija's [End Page 111] contribution is an effort at positive theory. By contrast, I aim to assess tax incentives normatively.

It is useful to recall that the practice of philanthropy is ancient. What is novel about the contemporary practice of philanthropy is the availability of tax incentives. The charitable contributions deduction in the United States is less than 100 years old, created by the US Congress in 1917 shortly after the institution of a system of federal income taxation in 1913. Similar incentives built into tax systems exist in most developed and many developing democracies.

Because the tax incentive constitutes a subsidy—the loss of federal tax revenue—it is no exaggeration to say that the United States and other countries currently subsidize the liberty of people to give money away, foregoing tax revenue for an activity that for millennia has gone unsubsidized by the state. The United States has the most generous subsidy structure. Charitable giving in 2012 was just over $315 billion and cost the US Treasury more than $50 billion in lost tax revenue.

I focus special attention on the incentive mechanism currently used in the United States and in many other countries: the charitable contributions deduction, a deduction of charitable gifts from a citizen's taxable income. Most countries use some kind of deduction scheme, including Australia, Germany, Japan, France, India, Spain, South Africa, Egypt, Mexico, the Netherlands, Russia, and Thailand. To the best of my knowledge, only Sweden provides no subsidy structure at all for charitable giving. The mechanism of an income tax deduction for a charitable donation works by creating a subsidy at the rate at which the donor is taxed. So a person who occupies the top tax bracket—in 2011, 35 percent in the United States—would find that a $1,000 donation actually "cost" her only $650. The government effectively pays $350 of her donation, subtracting this amount from her tax burden. Similar incentives exist for the creation of private and family foundations, and for contributions to community foundations, where donations and bequests to a foundation are deducted from estate and gift taxation. In permitting these tax incentives, federal and [End Page 112] state treasuries forego tax revenue. Had there been no tax deduction on the $1,000 contribution, the state would have collected $350 in tax revenue.

The remainder of this essay identifies and assesses various formulations of the most common justification for instituting tax incentives for charitable giving. That justification—which I'll call the subsidy rationale—is that the deduction efficiently stimulates the production of public goods and services. I do not believe that the current form of the subsidy—the charitable contributions deduction—can succeed. In conclusion, I gesture toward an alternative and potentially more compelling justification for tax incentives for charitable giving, a pluralism rationale. This alternative justification links the incentive to the effort to decentralize authority, to some degree, in the production of public goods and, in the process, to support a pluralistic civil society in a flourishing democracy. My references here are chiefly to the practice and regulatory framework of philanthropy in the United States, though I believe my analysis holds more generally for any liberal democracy (I draw heavily in what follows from portions of my essay "Toward a Political Theory of Philanthropy"; see Reich 2010).

THE SUBSIDY RATIONALE

The most common defense of the charitable contributions deduction is that the state accomplishes something of important social value by providing subsidies for people to be charitable. It is justifiable for a democratic government to provide incentives for charity because it is believed that the incentives stimulate the production of something of greater social value than what the state could have produced on its own had it not offered the incentives. Such a justification is assumed most frequently by economists, such as Bakija (2013), when they study the elasticity of the incentive.

The incentive is a tax expenditure, the fiscal equivalent of a direct spending program (on the concept of a tax expenditure, see Surrey and McDaniel 1985). When the state allows citizens to deduct [End Page 113] their charitable contributions from their taxable income, the state foregoes tax revenue, which is to say that all taxpayers are affected. They are affected in (at least) two important ways. First, they stand to lose some portion of the benefit they receive from direct government expenditures. If every citizen gains some fraction of the total revenue of the federal budget, the loss of billions of dollars in tax revenue through the deduction lowers every citizen's fractional benefit. Second, citizens lose in democratic accountability, for the foregone funds are not accountable, or even traceable, in the way that direct government expenditures are. To give an obvious example: citizens can unelect their representatives if they are dissatisfied with the spending programs of the state; philanthropic foundations also have domestic and global spending programs, partly supported through tax subsidies but their directors and trustees cannot be unelected. Thus the success of the subsidy rationale depends on whether the benefits brought about by the subsidy exceed the costs of the lost tax revenue.

The subsidy rationale has been invoked in several US Supreme Court decisions. "Both tax exemptions and tax deductibility are a form of subsidy that is administered through the tax system," the court noted in 1983 in Regan v. Taxation with Representation (461 US 540, 1983, 544). The deduction is also included in the annual federal tax expenditure budget issued by the US government.

The subsidy rationale is not concerned with donors per se but with the outcome of donations. Even so, the particular vehicle used in the United States to provide the subsidy—a deduction from taxable income—is vulnerable to powerful criticisms that call attention to fair treatment of donors. First, the deduction is available only to itemizing taxpayers, a group that constitutes roughly 30 percent of all tax returns. The other 70 percent of taxpayers, though they may make substantial charitable contributions (as an absolute sum or percentage of income), are excluded because they take the standard deduction. Thus the subsidy is capricious, for its availability depends on a characteristic, one's status as an itemizer, that has nothing whatsoever [End Page 114] to do with the value of giving. If the subsidy is justified because it produces some social good, then why should two donors who make identical donations to identical organizations, ostensibly producing the identical social good, be treated differently by the tax code?

Second, in a system of progressive taxation the deduction is tied by definition to marginal tax brackets. The richer you are, the less a charitable contribution actually costs you. The deduction functions as an increasingly greater subsidy and incentive with every higher step in the income tax bracket. Those at the highest tax bracket (35 percent in the US in 2011) receive the largest deduction, those in the lowest tax bracket (10 percent) receive the lowest deduction. Scholars have dubbed this the "upside-down effect," the result of which is that, for charitable deductions, "the opportunity cost of virtue falls as one moves up the income scale" (Musgrave and Musgrave 1984, 348).

According to a recent report by the Congressional Budget Office, the top income quintile (those earning more than $162,000 in a year) claimed almost half of all tax expenditures (see Distribution of Major Tax Expenditures in the Individual Income Tax System, https://www.cbo.gov/publication/43768). The tax benefit of the charitable contributions deduction to the wealthiest quintile is starker still: of the nearly $39 billion of tax expenditures for individual contributions, fully $33 billion was claimed by the wealthiest quintile.

When we combine this disparity in benefit with the fact, documented in the earlier part of this essay, that the giving preferences of the wealthy are the least likely to be targeted toward support for the poor, we realize that the subsidy for charitable giving is steeply regressive. The subsidy serves mainly to support the giving of the wealthy, and the giving of the wealthy rarely is directed at support of the needy.

These criticisms are not, however, criticisms of the subsidy rationale as such. They are criticisms of the mechanism for delivering the subsidy—currently in use in the United States and in many other countries—the income tax deduction. Reform of the subsidy mechanism could eliminate or mitigate the problem. The deduction could [End Page 115] be extended to all taxpayers regardless of itemizer status; or the deduction could be eliminated in favor of a partial or total tax credit; or the incentive could set a floor of initial giving—say 2 percent of income—before taxpayers would become eligible for a deduction. This latter mechanism would steer the tax subsidy in the direction of the poor, for the poor give away a higher percentage of their income than all other income classes (possibly excepting the wealthiest 1 percent) (see James and Sharpe 2007 for more on U-shaped charitable giving patterns).

How then might we assess the subsidy rationale as a whole? One obvious way to evaluate the subsidy rationale rather than just the subsidy mechanism currently in use is to look to the social good the subsidy produces and the efficiency with which it is produced.

Supposing that the goods produced by charitable recipients were of social benefit, we might ask, for instance, whether the subsidy is "treasury efficient." Does the subsidy shake off more in donations than it costs in federal tax revenue? If so, the subsidy is treasury efficient. Economists will then argue about the optimal rate of the subsidy or how to stimulate the most giving at the least cost to the treasury. Empirical analyses of the tax deduction in the United States show that the deduction is indeed treasury efficient, though significantly less so than initially was thought (see Bakija 2013).

While treasury efficiency assures us that the subsidy is not a mere reward for charitable giving that would occur even in the absence of the subsidy—a loss of tax revenue to produce something that would occur anyway—whether it justifies the subsidy depends on the initial supposition that the goods produced by charitable recipients are of broad social value.

When we inquire into the social good produced by charitable donations, rather than focusing squarely on questions of treasury efficiency, two problems present themselves, at least in the US context.

First, as mentioned earlier, United States law permits a truly kaleidoscopic landscape of public charities to receive tax-deductible charitable contributions.² Some and perhaps many of the social goods [End Page 116] produced by charities will be of no value whatsoever to certain citizens. Because churches are eligible to give tax deductions to donors (for example, congregants) for contributions, atheists are vicarious donors to churches through the tax subsidy. By contrast, Catholics are vicarious donors to Planned Parenthood and its support for abortion rights. Such examples are easily multiplied. The basic point is that the subsidy cannot be justified as a Pareto improvement, where some benefit and no one is made worse off (this is Gergen's argument; see Gergen 1988, 56). At best, the subsidy is a so-called Kaldor-Hicks improvement, where the gains for those who consume the particular social good produced by charity offset the losses to those with no interest in that social good.

But relying on a Kaldor-Hicks improvement as the standard for justifying the subsidy rationale raises a second set of problems. For obvious reasons, the beneficiaries of a deduction are highly skewed toward upper-income earners. Wealthier individuals donate more as an absolute amount (but not as a percentage of income) and receive a larger subsidy for giving (the upside-down effect) and thus claim a large share of the deduction. The result is a plutocratic bias in the subsidy, where the favored beneficiaries of the wealthy receive the lion's share of the subsidy.

The plutocratic bias is troubling, for systematic overattention to the interests and preferences of the wealthy against the interests and preferences of the middle-class and poor seems a strange, indeed unjust, basis for social policy. But the trouble might be undercut if the product of charitable giving were pure public goods, in the economic sense—namely, goods that are nonexcludable and nonrivalrous. If wealthy people donate to create goods that no one can be prevented from enjoying and where one person's consumption does not reduce the amount available to others, then the plutocratic bias nevertheless redounds to the advantage of all citizens.

But our earlier examination of the distribution of charitable dollars shows how few public charities produce pure public goods. Hospitals and universities, for instance, together account for more [End Page 117] than half of the revenue of all nonprofit organizations in the United States. Both hospitals and universities can easily exclude persons who cannot pay for their services.

Leaving aside the strict conditions of pure public goods, the concern about plutocratic bias might be mitigated if the favored beneficiaries of the charitable givers, and the wealthy especially, were charities engaged in social welfare or services for the poor. At the very least, then, the effect of charitable giving would be to some degree redistributive. Yet we have seen that this is not the case, at least in the United States.

Suppose now that charitable donations were redistributive in the sense that gifts from the relatively wealthy flowed to the relatively poor. Granting this, we may nevertheless not yet conclude that nonprofit organizations and foundations are in fact redistributive because we must still account for the tax concessions to philanthropy and the counterfactual scenario in which the money flowing into nonprofit organizations and foundations would have been taxed and become public revenue. The relevant question is not merely, "Is philanthropy redistributive?" but instead, "Do philanthropic dollars flow more sharply downward than government spending does?" In order for the return, so to speak, on the public's investment in philanthropy to be worthwhile, philanthropy must do better than the state would do had it taxed the philanthropic assets.

Answering this counterfactual question is difficult. We are forced to speculate about how the state might spend the tax revenue it could have collected if it hadn't extended the tax concessions to philanthropists for their gifts.³ I will not make any such speculation here. Instead, I wish to note that anyone who seeks to ground the special tax treatment of philanthropy in the United States on the sector's redistributive outcomes must confront at least three reasons to be suspicious that any such redistribution actually occurs. There is the first, obvious difficulty that a motley assortment of nonprofit groups all qualify for 501(c)(3) status, puppet theaters and soup kitchens alike. There is the second difficulty that religious groups dominate as the beneficiaries of individual charitable dollars. And there is the [End Page 118] third difficulty that the burden on the sector's advocate is to show not merely that philanthropy is redistributive but that it is more redistributive in its actions than would be the government. In short, we have some good prima facie reasons to doubt that philanthropy is redistributive in effect.

These problems once again target the mechanism in the United States and elsewhere to deliver the subsidy: the tax deduction. The plutocratic bias in the subsidy and the lack of redistribution could be altered by changing both the mechanism of the subsidy (change to a capped tax credit, for instance) and limiting the kinds of organizations that are permitted to receive tax-deductible donations (eliminating churches and cultural organizations, for instance). Whatever the remedy, the expectation would be that the subsidy must still be efficient. To be justified, the subsidy must cost less to the treasury than it produces in social benefits.

I shall not explore these sorts of remedies here. Instead, I underscore the argumentative burden of the advocate for a subsidy rationale. The advocate needs to meet the following objections. The current mechanism in use in the United States and many other countries—the charitable contributions income tax deduction—suffers from various shortcomings, most prominently its capricious exclusion of nonitemizers and its upside-down subsidy structure. The number and wildly diverse types of eligible organizations for receiving tax deductible gifts mean that few charitable organizations produce anything like pure public goods. And supposing that charitable giving preferences were at some point to come to favor caring for the needs of strangers, the subsidy rationale defender would still need to show philanthropic giving was more redistributive than would be government spending had it instead collected the foregone tax expenditures.

Economists show us that the charitable tax deduction does in fact stimulate more charitable giving than would occur in the absence of the deduction. Just how much more is a matter of some controversy, and just how sensitive various donors are to the subsidy is an open question.

Were the deduction merely a reward for giving, then it would be harder still to defend it. But even when we accept the finding, as [End Page 119] in Bakija's 2013 paper, that the incentive is more than just a reward, powerful objections to the subsidy remain. I conclude that the subsidy rationale for justifying tax incentives for charitable giving is not compelling, all the more so if one believes that charity ought to be mainly an activity that serves to benefit the poor.

CONCLUSION

But perhaps we should not view the purpose of charity as an activity intended to support the basic needs of others. If charity can be construed as having a broader scope, then an alternative rationale in support of tax incentives for charitable behavior may be available. I conclude by gesturing toward such an alternative rationale. (I provide more detailed argument for an alternative rationale in Reich 2010.)

The alternative is what I would call the pluralism rationale. The pluralism rationale comes in several stripes and cannot be called a unified theory. The basic idea is that the tax incentive to make charitable donations should not be justified on the basis of assessing the discrete social goods or outputs of the various nonprofit organizations funded through these donations. Instead, the tax incentive is justified for its role in stimulating or enhancing the voice of citizens in the production of a diverse, decentralized, and pluralistic associational sector, which is in turn thought to be a bedrock of a flourishing liberal democracy. If nonprofit organizations are the institutional face of associational life, then stimulating charitable donations to a wide array of nonprofits might amplify the voice of citizens and enhance civil society to the overall benefit of liberal democracy. Rather than focus on the matrix of goods produced by charitable organizations, the focus here is on the creation and sustenance of a diverse slate of organizations themselves. The public good or social benefit being produced is associational life itself, not the catalogue of public goods or benefits produced by the roster of organizations that constitute civil society.

Note that this is still a subsidy theory but there is no necessary demand that the subsidy be treasury-efficient. Even if there is a net [End Page 120] loss to the treasury in the production of the social goods generated by nonprofit organizations—if the state could more efficiently deliver these goods itself—the pluralism rationale holds that the subsidy is nevertheless worthwhile. Of course, there is no bias against the efficient production of goods, but the pluralism rationale does not demand efficiency for the success of the argument. The state might justifiably forego tax revenue for the sake of fostering citizens' voices and the sustenance of a diverse associational sector.

The idea behind the pluralism rationale is captured in a US Supreme Court opinion from Justice Lewis Powell, where he takes issue with the notion that the purpose of the nonprofit sector is efficiently to deliver or supplement services or social goods that the government would otherwise supply through direct expenditures. Powell rejects the view that the primary function of a tax-exempt organization is to act on behalf of the government in carrying out government-approved policies. In my opinion, such a view of 501(c)(3) ignores the important role played by tax exemptions in encouraging diverse, indeed often sharply conflicting, activities and viewpoints. As Justice William Brennan has observed, private, nonprofit groups receive tax exemptions because "each group contributes to the diversity of association, viewpoint, and enterprise essential to a vigorous, pluralistic society." Far from representing an effort to reinforce any perceived "common community conscience," the provision of tax exemptions to nonprofit groups is one indispensable means of limiting the influence of governmental orthodoxy on important areas of community life (Bob Jones University v. United States 461 US 574, 1983).⁴

The pluralism rationale rejects the idea that the sole or primary aim of charity is to care for the needs of strangers. It finds nothing inherently objectionable in the pattern of distribution canvassed earlier.

Though people have engaged in philanthropy for millennia, the practice of giving money away has only recently become a tax-subsidized activity. Philanthropy is now embedded within a framework of public policies, usually centered on the tax regime, that structure its [End Page 121] practice and alter its shape from what it would otherwise be without the state's intervention. Though nearly all liberal democracies have tax incentives for charitable donations, the justification for this practice is not well understood or theorized. I have examined the most common justification for providing tax incentives for philanthropy, a subsidy rationale, and I have mentioned a potential alternative, a pluralism rationale. Neither justification, however, provides support for the actual design of most tax-subsidized giving, where a wide array of eligible recipient organizations and a tax deduction for giving are the favored mechanisms. In my view, then, a philosophical analysis of philanthropy might offer a defense, or several distinct defenses, of state incentives for giving money away, but the current practice of state-supported philanthropy, especially in the United States, is indefensible.