(Illustration by Doug Chayka) 

We live in a second gilded age of massive and still growing inequality. While this is a foe to civic comity, it is a friend to private philanthropy. In the United States, there were approximately 200 private foundations in 1930 possessing aggregate assets of less than $1 billion. In 1959, there were more than 2,000; in 1985, more than 30,000 private foundations. As of 2014, the number was nearly 100,000, with total capitalization of more than $800 billion. What Andrew Carnegie and John D. Rockefeller were to the early 20th century, Bill Gates and Mark Zuckerberg are to the 21st. With every passing year, a new billionaire appears to arrive on the philanthropic scene, declaring an intention to make the world a better place. Amazon founder Jeff Bezos is now the world’s richest man, at $160 billion. His September 2018 announcement that he would donate $2 billion to combat homelessness and create a network of preschools is only the most recent example, and this is likely just his initial foray into big philanthropy.

The scope of philanthropy goes far beyond billionaires and their foundations, though. Despite the eye-popping size of large foundations and the growth in the total number of foundations, the overwhelming majority of total giving, at least in the United States, comes from living donors making charitable contributions. Americans donated more than $410 billion to eligible nonprofit organizations in 2017. Of that total, giving by living individuals accounted for $287 billion, or 70 percent. Nearly all Americans donate some amount of money every year. A small donor does not wield the same kind of power as does a big philanthropist. Yet the distribution of small giving matters a great deal in the aggregate, fueling the operation of thousands of nonprofit organizations, and small donors enjoy the same discretion as big philanthropists and also can benefit from tax incentives for their giving. Any consideration of philanthropy must go beyond the Rockefellers and Gateses of the world and attend to the amount and significance in a democratic society of ordinary charitable giving.

It may seem that philanthropy is just voluntary activity, a result of the exercise of individual liberty. But as I argue in my new book, Just Giving, this is a mistake. It is indeed voluntary, but because philanthropy is a tax-subsidized activity, it is partly paid for by all taxpayers. Strictly speaking, then, donors are not exercising a liberty to give their money away; they are subsidized to exercise a liberty they already possess. Unlike the Carnegie-and-Rockefeller era, when enormous philanthropic entities were created without any tax concessions for doing so (because the personal income taxation had yet to be adopted; it would arrive only in 1917), today philanthropy is partially underwritten by the state through a complex web of advantageous tax laws that apply both to donors and to nonprofit organizations and private foundations. In the United States, subsidies for charitable contributions cost citizens more than $50 billion in forgone federal tax revenue in 2016.

Contemporary philanthropy in democratic societies is embedded within a set of legal rules that structure and encourage it. Whether, when, to whom, and how much people give is partly a product of laws that govern the creation of nonprofit organizations, charitable trusts, private and community foundations, and so on, and spell out the rules under which these may operate. These factors are shaped by tax policies that set up special exemptions for philanthropic and nonprofit organizations and that frequently permit tax exemptions for individual and corporate donations of money, property, and other assets. They are governed by laws that enforce donor intent, often beyond the grave, creating philanthropic projects and entities that can exist, in principle, in perpetuity. These governance arrangements are an essential component of the modern practice of the time-immemorial activity of giving.

The policies that structure American philanthropy are broken. There’s a long list of reasons why this is so. Donor-advised funds are spreading like kudzu, increasingly dominating the list of most popular charitable causes and, in the process, warehousing increasing sums of philanthropic wealth while donors take immediate advantage of tax benefits for giving. President Donald Trump’s 2017 Tax Cuts and Jobs Act significantly diminished the incentive for giving by capping total itemized deductions and raising the standard deduction. Numerous studies predict a decrease in charitable giving in 2018. And our wealthiest donors are making philanthropy into a political weapon, funneling dark money through social welfare organizations or 501(c)(4)s and, like Mark Zuckerberg, Pierre Omidyar, and Laurene Powell Jobs, setting up limited liability companies (LLCs) at least partly in order to avoid the transparency requirements that attach to foundations.

But, worse than being ineffective and broken, the policies that structure American philanthropy are also indefensible. The array of policies designed to stimulate the charitable donations of ordinary citizens and the philanthropic projects of the wealthy—chiefly through private foundations—subvert, rather than support, democratic aims. Philanthropy too often undermines democracy, and it is our policies—not the preferences of individual donors or operations of particular nonprofits—that are largely to blame.

How can philanthropy support democracy? To answer this question, we need to operate on two levels. We need to target and address the injustice at the heart of the most important and most common policy instrument at use in the United States, and in many other countries, concerning philanthropy: the tax deductibility of charitable contributions. Deploying tax concessions in the form of tax-deductible contributions cannot be defended. We should replace tax deductibility with a flat tax credit for donors. And we need to recognize that even if there were no tax advantages at all, the ultrawealthy would still have enormous power. Big philanthropy, whether tax subsidized or not, is an exercise of power—the attempt to direct private assets toward some public purpose. It is a form of power that is unaccountable, low on transparency, donor directed, and by default perpetual. Big philanthropy is a plutocratic element in democratic society. The challenge is to craft, through various policies and social norms, a framework that domesticates plutocrats to serve democratic ends.

The Plutocratic Bias of Tax Deductibility

Few things are more soporific than analyzing tax policy. And yet the subject is of enormous consequence in philanthropy. It’s where much of the governance and regulation of philanthropy rest, and where the tax treatment of nonprofit organizations and charitable contributions set the incentive structure for giving. Philanthropy would not disappear if tax incentives for giving were eliminated, but the total amount and overall distribution of philanthropy would likely be dramatically different without the incentives.

The policy instrument of choice in contemporary philanthropy is the tax deduction. There are two big problems with this: A deduction for charitable contributions rewards donors arbitrarily—treating differently two donors who make identical contributions to the identical organizations, ostensibly producing the identical public benefit—and it systematically benefits the wealthy, amplifying their voices and giving preferences over everyone else.

First introduced into the tax code in the War Revenue Act of 1917, the tax deduction allows individuals to deduct the sum of all eligible charitable donations from their taxable income. Over the course of the past hundred years, Congress has often modified the provision, changing and occasionally eliminating the ceiling on total charitable deductions and expanding the set of eligible recipient organizations of tax-deductible gifts. But the heart of the policy has always been the same: a deduction of charitable giving from taxable income. In some form or another, the deduction applies to contributions to public charities, donor-advised funds, private and family foundations, and community foundations, and to charitable bequests.

Let’s consider this from the perspective of two would-be donors. Take Annie, who rents an apartment and brings home the median personal income in 2017, roughly $31,000. And take Bill, who owns a house and brings home a personal income in the top 1 percent, roughly $300,000. Assume that Annie and Bill both wish to make a $1,000 donation to their local food bank.

A tax deduction for a donation creates a subsidy by the government at the rate at which the donor is taxed. Progressive taxation will levy different taxes on people with higher and lower incomes. In 2017, Annie falls into the 15 percent tax bracket. A $1,000 donation to the food bank would diminish her taxable income by $1,000. As a result, Annie would find that her $1,000 donation cost her $850, because the government would effectively pay 15 percent, or $150, of her donation, subtracting this amount from her tax burden.

By contrast, Bill occupies the top tax bracket—39 percent in 2017—and would find that a $1,000 donation actually cost him only $610. The government would effectively pay $390 of his donation.

In extending these tax incentives, federal and state treasuries forgo tax revenue. Had there been no tax deduction on Bill’s $1,000 contribution, the state would have collected an additional $390 in tax revenue. Or, to put it differently, tax incentives for philanthropy constitute a kind of spending program or “tax expenditure.” Tax incentives for philanthropy are one of the largest tax expenditures for individuals in the US tax code, and they amount to massive federal and state subsidies for the operation of philanthropic and charitable organizations and to the individuals and corporations that make donations of money and property to them.

But notice how the policy instrument treats Annie and Bill differently. They make identical donations to the identical organization, and yet, despite his higher income, Bill receives a larger subsidy than Annie. Annie’s $1,000 donation costs more than Bill’s $1,000 donation—$850 for Annie and $610 for Bill. This is known as the “upside-down effect” of tax deductions, where the deduction functions as an increasingly greater subsidy with every higher step in the income tax bracket.

Source: "Patterns of Household Charitable Giving by Income Group, 2005," The Center on Philanthropy at Indiana University, Summer 2007. 

Since the same social good is ostensibly produced in both cases—the food bank receives $1,000 from each—the differential treatment appears totally arbitrary at best and unfair at worst. If anything, lower income earners might seem to warrant the larger subsidy in order to lower the cost of their charitable giving; in light of the declining marginal utility of additional dollars for people at the top of the income scale, they can afford a “higher price” for charitable donations than can poor people. The upside-down phenomenon is not specific to the tax deduction for charitable donations, of course. Deductions in general massively favor the wealthy. In 2013, the wealthiest decile of earners claimed more than two-thirds of all tax deductions.

Finally, and perhaps most glaringly indefensible, the tax subsidy for charitable contributions is available only to those individuals who itemize their deductions—people who opt not to take the so-called standard deduction on their income tax. This effectively penalizes, or fails to reward and provide an incentive for, all people who do not itemize their deductions, a group estimated to be roughly 90 percent of all taxpayers after the 2017 Trump tax reform takes effect. In this respect, only the very wealthy receive any tax benefit from the charitable deduction. And yet almost all Americans make annual charitable contributions.

The upshot is that the charitable giving policies in the tax code are deeply inegalitarian: They systematically favor the rich in providing them with larger benefits. It’s of course true that wealthy people give away more money in absolute terms than do poor people. But why should public policy differentially reward the rich over the poor? Why should more than two-thirds of the tax expenditures for charitable giving be attached to the giving preferences of the wealthiest 10 percent of Americans? The relevant issue here, therefore, is not just that the incentive applies unequally to donors of different tax-filing statuses and income levels; it’s that the public funds forgone in the tax deduction are flowing disproportionately to the favored charitable organizations of the rich. Tax policy in the realm of charity favors the wealthy and, by extension, weights the preferences of the wealthy over those of the poor in the nonprofit organizations they fund. The 1 percent receive a tax-policy megaphone and use it to promote causes very different from those favored by the middle class and poor. (See “The Very Rich Give Differently” on this page.)

In a democracy, the justification of any tax incentive for donations must be rooted in something more than the desire to reward people for practicing charity. I believe that the best justification for tax-subsidized giving is that charity is essential to the project of supporting civil society. A tax incentive is justified for its role in stimulating or amplifying the voice of citizens in the production of a diverse, decentralized, and pluralistic associational sector, which is itself important because it is considered a bedrock of a flourishing democracy. If nonprofit organizations constitute, to a significant degree, the institutional matrix 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.

But if pluralism in civil society forms a foundational basis for the policies that structure charitable giving, a plutocratic bias in the policy instrument is unjustifiable. With the upside-down subsidy and the capricious exclusion of nonitemizers in the current policy scheme, we get not equal citizen voice in civil society but plutocratic citizen voice, underwritten and promoted by tax policy.

In theory, it would be quite simple to remedy these problems. Tax policy could allow nonitemizers to deduct their charitable contributions from their income (on top of the standard deduction). Better, since this solution would still leave the upside-down effect in place, policies could allow all donors an identical nonrefundable and capped tax credit, rather than a tax deduction, for donations.

By offering an equivalent tax credit to all donors (say, 25 percent of any donation) and capping the total annual credit at some level (say, $1,000), the fix avoids the upside-down structure of the deduction, offers an equal credit to all donors, and affords donors the liberty to continue to give money away after the cap has been reached, but no longer with any state subsidy to do so. The policy proposal bears a resemblance to a stakeholding grant or a campaign finance voucher scheme for each citizen, though, rather than directing the use of the stakeholding grant for investment in one’s own projects or a voucher for expressing political voice, the tax credit could be directed only toward eligible civil society organizations. Call it a civil society stakeholding grant, assigned on an equal basis to every citizen in the form of a nonrefundable tax credit, with Bill Gates receiving the same-size credit as every other citizen.

The credit could even be designed to try to surmount one of the most stubborn and yet unfamiliar features of charitable giving: the fact that American giving has hovered around 2 percent of gross domestic product for several decades. What might boost charitable giving above that rate? Rather than constructing a tax credit as a percentage of any charitable donation, eligibility for the credit could be conditional on first giving away 3 percent of one’s income without any tax advantage for doing so. After donating 3 percent, a person would receive a civil society tax credit of, say, $1,000 to direct to the charities of her choice. If every person knew that by giving away 3 percent of income they would receive a $1,000 credit for further donation, that might induce higher rates of giving.

The Plutocratic Power of Foundations

The tax deduction contains a plutocratic bias, favoring the wealthy and their charitable projects. A tax credit would correct this bias, treating donors equally by granting to each donor an equal credit. But even with a tax credit, the wealthy would still have greater power than poorer people, simply by virtue of their greater resources. Indeed, even if there were no tax incentive whatsoever for charitable giving, the ultrawealthy, through their greater giving capacity, would exercise more power than others. Bill Gates and I may receive an identical tax credit, and the policy instrument may treat us without plutocratic bias. But, let’s face it, Bill Gates will have a greater effect on civil society than I will.

In understanding why the policies that structure American philanthropy are indefensible, we have to move to a second level of analysis, from plutocratic bias in tax-deduction policy to the plutocratic power that the wealthy exercise. It’s not the tax deduction that’s the problem here; it’s the structure of the private foundation. We need to confront the largely unaccountable, nontransparent, donor-directed, tax-advantaged, and by default perpetual power of big philanthropy. And we need to ask whether it is healthy for democracy that the number of foundations in the United States has exploded over the past few decades.

Though rooted in historical traditions, the modern private foundation in the United States is a creation of the age of Carnegie and Rockefeller. The idea behind the Rockefeller Foundation and the similarly minded Carnegie Corporation was to establish an entity with broad and general purposes, intended to support other institutions and indeed to create and fund new organizations, seeking to address root causes of social problems, rather than deliver direct services (work “wholesale,” not “retail”), and designed to be administered by private, self-governing trustees, with paid professional staff, acting on behalf of a public mission.

Private foundations are, more or less by definition, the legal sanctioning—or, more precisely, the legal promotion—of plutocratic voices in democratic societies. This concept was recognized as such in the Carnegie-and-Rockefeller era. When Rockefeller came before the US Congress to seek a federal charter to incorporate the Rockefeller Foundation, he encountered widespread criticism. Louis Brandeis, the trust-busting “people’s attorney” who would later become a US Supreme Court justice, testified before the Senate Industrial Relations Committee in 1916 that the Rockefeller Foundation was “inconsistent with our democratic aspirations” and confessed to “grave apprehensions” about the power lodged in the hands of a few wealthy men.

Democratic societies are committed to much more than a representative government with free and fair elections. They are also committed to the equal standing of citizens and an equal respect for their interests. Such equal standing and respect are manifest when citizens are formally equal under the law—there is no second-class citizenship—and when all citizens possess an equal opportunity for political influence and participation.

This shared expectation of political equality sits in tension with the existence and growing power of private foundations to influence public policy. The larger the foundation, the greater the potential power. Think here of Bill Gates, whose philanthropy permits him to stride upon the world stage as if he were a head of state. Why should we grant such an outsize voice to any citizen of our democracy? Can we publicly justify our current laws that define how foundations may be created and structure how they operate? Perhaps foundations play salutary roles in democratic societies, despite being exercises of unequal power and expressions of plutocratic voice—or could play such a role if they were subject to different legal arrangements.

For most of the 19th century, creating a grantmaking foundation at one’s private initiative with one’s private wealth was not permissible; authorization and incorporation by a democratic body were necessary. The prospect that general-purpose foundations might be brought into existence was viewed as a threat to democracy. Even after such foundations were created, they continued to be treated with public scorn and skepticism. In a decision that seems positively unimaginable from today’s vantage point, the regents of the University of Wisconsin passed a resolution in 1925 that banned the university from accepting philanthropic donations from foundations.

We have come a long way since then. Philanthropists are today widely admired, their names trumpeted from buildings and their photos gracing magazine covers. The permission to create a foundation, moreover, is both freestanding—not requiring approval by a democratically elected body—and, as with ordinary charitable donations, subsidized with tax advantages.

But the pendulum has swung too far. I believe we need policies and social norms that render private foundations supportive, not subversive, of democratic aims. To understand how, we must first understand the institutional oddity that is the private foundation.

Foundations Lack Accountability

In the commercial marketplace, if a company fails to make a profit because consumers opt not to purchase the goods it sells, the company goes out of business. In the public institutions of a democratic state, officials responsible for crafting law and allocating tax dollars must stand for election; if citizens do not approve of the public policies and spending decisions of their representatives, they vote for replacements at the next regularly scheduled election.

By contrast, foundations have no market accountability; they have neither goods for sale nor marketplace competitors. Instead of selling anything, foundations give money away to other organizations, whose own livelihood frequently depends on continuing support from foundations. Foundations have no consumers or competitors, only supplicants for their money, in the form of grants. If citizens do not like a foundation’s grantmaking decisions, they have no recourse, because there is nothing to buy and no investors are holding the foundation accountable.

Moreover, foundations have no electoral accountability; no one in a foundation stands for election, regardless of what the public thinks about the distribution of its grants. Suppose a group of people disapproves of what the Gates Foundation, or any other foundation, is doing. What then? There’s no way to unelect Bill and Melinda Gates. Referring to the foundation’s education grantmaking, critic Diane Ravitch has called Bill Gates the nation’s unelected school superintendent.

Compounding the lack of any formal accountability is the difficulty any foundation has in developing informal processes to generate honest feedback from grantees, beneficiaries, and the general public. People who interact with foundations are typically deferential and solicitous, pleading for a grant or seeking the next grant. There is little incentive for a potential or actual grantee to offer critical feedback to a foundation. Every person who works in a foundation understands what comes with the territory: People who become foundation officers are transformed overnight into the smartest and best-looking people in the room.

Foundations Lack Transparency

Compounding the accountability problem, foundations are frequently opaque, drawing blackout shades across their windows. They face a legal requirement to pay out 5 percent of their assets every year and file an annual tax form with some basic data. But they need not have an office, a telephone number, or a website. (Fewer than 10 percent of foundations have a website, according to the Foundation Center.) They need not publish an annual or quarterly report or articulate any grant-
making strategy. They need not evaluate their grantmaking; if they do, they need not make such evaluations public. They need not report on trustee decision making.

Some foundations—especially the largest and most professionalized—do operate transparently, providing all of the above information and more. But this is a function of the idiosyncratic preference of a particular foundation, not a legal requirement or professional norm. A great many small family foundations operate with virtually no public trace, save their legally required annual tax filing. It’s not just small family foundations that seek to avoid transparency, either. In 2017, the Paradise Papers, a leak of documents to the Süddeutsche Zeitung, revealed that James Simons, a billionaire hedge fund manager, had created the Simons Foundation International with an estimated $8 billion endowment. Incorporated in Bermuda, its assets thus entirely tax-free, the foundation had, according to a 2017 profile of Simons in The New Yorker, no Web page or public presence at all.

Foundations are legally designed to enshrine donor intent and protect philanthropic assets in perpetuity. Thus does the dead hand of the donor potentially extend from beyond the grave to strangle future generations. Foundations must be governed by a board of trustees, but the donors and their family or trusted associates can serve in this role; there is no requirement of community or public governance. Wealth management firms routinely market their services in setting up a family foundation as vehicles for the intergenerational transmission and sustenance of family values. A founding donor may thereby control the governance and purpose of a foundation forever.

For foundations with few or no formal accountability mechanisms, practically no transparency obligations, a legal framework designed to honor donor intent in perpetuity, and generous tax breaks to subsidize the creation of a foundation, what gives them their legitimacy in a democratic society?

The Discovery Case for Foundations

The positive case for foundations depends on changing the policies that govern them and creating new social norms that will influence wealthy donors. And it depends on transforming the apparent vice of unaccountability into a virtue. Because of their size and longevity, foundations can operate on a longer time horizon than can businesses in the marketplace and elected officials in public institutions, and can take risks in social policy experimentation and innovation that we should not routinely expect to see in commercial firms or state agencies. I call this the discovery argument on behalf of foundations.

Begin with an uncontroversial supposition: A democratic state wishes to advance general welfare or to pursue the aims of justice, however understood. But democratic representatives do not know the best means for achieving such aims, either at any given moment or, especially, with an eye toward changing social conditions in the future. What kinds of policies and programs, for instance, will best promote educational opportunity and achievement? Some believe universal preschool is the answer, others a better school finance system, others improved and more pervasive opportunities for online learning.

My aim here is not to defend foundations,
but to identify the right standard by which
to assess them.

To answer such questions, a democratic society, recognizing that its elected leaders are not all-knowing, that reasonable disagreement on the best means to pursue just ends is likely, and that social conditions are always evolving, might decentralize experimentation in social policy so that it can identify and adopt better and more effective policies at realizing democratically agreed-upon aims. Moreover, this need for experimentation is never-ending. In light of constant change in economic, cultural, technological, and generational conditions, the discovery process is, in ideal circumstances, cumulative, in contributing to society a storehouse of ideal, or simply very effective, ways to address different contexts and shifting priorities.

To be sure, a democratic government can stimulate some measure of experimentation and risk-taking innovation on its own. It can, for example, invest in basic research with uncertain outcomes by directing public funding to research universities. It can develop federal structures of government that treat jurisdictional subunits as sites of policy experimentation—hence Brandeis’s famous description of American states as laboratories of democracy. Democratic government has good reason to be experimentalist, to approach policy and institutional design as a form of problem solving.

Such approaches notwithstanding, political leaders would also be right to harbor some skepticism that democratic government is ideally suited to carry out such experimentation itself. For one thing, citizens in a system of democratic governance tend to expect and prize tested and reliable outcomes in public policy. Elected representatives who allocate public funds to chancy strategies for solving social problems—in the sense that the selected policy may fail in delivering any benefits at all—also run the risk of being punished at the ballot box. Furthermore, wasteful government spending tends to be deplored, and yet experimentation requires that some trials fail if the approach is to deserve the label “experimentation” in the first place.

What extragovernmental structures, then, can be designed to carry out decentralized experimentation? My claim is that foundations are one such vehicle for this important work of discovery and experimentation.

Foundations have a structural advantage over market and state actors in this discovery effort: a longer time horizon. Once more, the lack of accountability may be a surprising advantage. An essential feature of the discovery argument focuses on the ideal conditions for innovation and risk-taking. Unlike profit-driven businesses, foundations are not subject to quarterly or annual earnings reports, bottom-line balance sheets, or impatient investors or stockholders. Commercial entities in the marketplace do not have an incentive structure that systematically rewards high-risk, long-time-horizon experimentation; they need to show results in order to stay in business. Similarly, public officials in a democracy do not have an incentive structure that rewards high-risk, long-time-horizon experimentation; they must quickly show short-term results based on the expenditure of public dollars to stand a strong chance of reelection.

Precisely because of their lack of ordinary democratic accountability and legal permission to persist for decades, foundations can fund experiments and innovation whose payoff, if it comes, benefits future, rather than present, generations in the long run. Moreover, because the universe of private foundations is diverse and donor driven, different foundations are likely to experiment with different approaches, improving the chance that they will find effective or simply better social policies or solutions to social problems.

How are we to evaluate philanthropic discovery? And what mechanisms could disseminate or bring to scale successful experiments that are the product of foundation-fueled innovation? Failed innovations die, though society has presumably learned something from the failure. Other foundations may take up and modify the experiment and later generate positive results. Still other foundation projects succeed in showing positive effects. Ideally, foundation-funded experiments would be subject to demanding social science review, not anecdotal reports from the field. But from the perspective of a foundation, success in its philanthropic giving consists not in funding innovative and risky social policy experiments and then sustaining the most successful of them forever. Because the assets of the marketplace and the state dwarf the assets of even the largest foundations, success consists in seeing the successful or proven policy innovations that the commercial marketplace or the state brings to scale.

Thus, the proper attitude of foundations toward democratic government is one of humble servant, instead of “smarter sector” or superior provider of social goods. A foundation project that was initially privately funded and democratically unaccountable auditions for adoption as a publicly funded and democratically accountable government responsibility.

The institutional design of foundations permits them to operate on a different time horizon than the marketplace and the government. Because their endowments are designed to last, foundations can fund higher-risk social policy experiments, and they can use their resources to identify and address potential social problems decades away or innovations whose success might be apparent only after a longer time horizon. In short, unlike business and the state, foundations can “go long.” They can be the seed capital behind one important discovery procedure for innovations in effective social policy in a democratic society.

Some of the greatest accomplishments of American foundations do seem to fit this model. Consider the quintessential example of successful foundation activity, Andrew Carnegie’s promotion of public libraries. Carnegie provided significant funding for the construction of libraries but conditioned his grants to municipalities on modest matching public dollars (usually 10 percent annually). Between 1911 and 1917, Carnegie’s philanthropy contributed to the creation of more than 1,500 public libraries. The library grant program was discontinued shortly thereafter, yet citizens found the libraries important enough to demand that they become the full responsibility of the local municipality. The privately financed public libraries successfully auditioned for inclusion in public budgets. Similar accounts could be given for other foundation successes, such as the development of Pell Grants in higher education and the coordination of a national 911 emergency response system.

Policy Recommendations

The discovery argument can be mobilized on behalf of some of the privileges that attach to contemporary foundations, and to some of their activities, but it has its limits. Namely, it does not justify the full range of legal permissions currently afforded to foundations. I am particularly skeptical that it is possible to defend the legal permission for a foundation to exist in perpetuity. I am also skeptical that the array of tax subsidies attached to philanthropy today is necessary for the creation and sustenance of foundations.

The discovery argument points the way toward some sensible policy recommendations for improving the work foundations should perform for democratic society. Three proposals specifically come to mind.

First, and perhaps counterintuitively, establish a floor, not a ceiling, on the size of foundations. The massive boom in small foundations is a problem. For foundations to be capable of providing sufficient risk capital for discovery, they must have significant assets and likely have a professional staff able to manage and disseminate their learning. By contrast, a small family foundation is not in a strong position to carry out such a task. The number of foundations with less than $1 million in assets nearly doubled from 1993 to 2013. Foundations with less than $1 million in assets rarely have a paid staff, almost never give away more than $50,000 in a year, and function more or less as a tax shelter and charitable checkbook for wealthy families. These families could accomplish the same outcome and produce the same public benefit by simply making an ordinary charitable donation, rather than setting up a foundation as the vehicle for their philanthropy. There is no good reason for the public to support, via tax benefits, the intergenerational transmission of family values by inviting family members to share in the governance, often with paid salaries, of a foundation that disburses less than $50,000 a year. And taxpayers would no longer be subsidizing enormous sums of money that have been committed to a foundation but have not yet been granted to charitable organizations. More than one-quarter of foundations’ total assets are held by just the 50 largest. What loss to public benefit would there be with a minimum asset threshold to create a foundation—say, $10 million or $50 million? I think very little, and quite possibly there would be some gain, for wealthy individuals under the minimum asset threshold might be more inclined to donate their money to public charities than to create their own family foundations.

Second, place time limits on foundations. Do we need the endowments that fuel foundation grantmaking to be perpetual? If so, do we need the founder’s intent to be honored in perpetuity? On this matter, I side with John Stuart Mill in believing that perpetuity is injurious to society. “There is no fact in history which posterity will find it more difficult to understand, than that the idea of perpetuity, and that of any of the contrivances of man, should have been coupled together in any sane mind,” Mill wrote in 1833. No argument on principle can specify the optimal life span of a foundation; the relevant consideration here is that it have an incentive structure that encourages work on a time horizon longer than that of other social institutions. What Rockefeller’s advisors proposed to Congress in 1911, that the life span of a foundation be capped at 100 years, or five generations, seems to me a more-than-adequate horizon in which to engage in the important, democracy-supporting work of discovery.

Third, apply the social norm of peer review to discovery. How might public policy or the creation of philanthropic norms guide private foundations and orient them more reliably toward the work of discovery? One possibility would be to expect in their annual public reporting a long-term or intergenerational impact statement. Foundations would submit to public scrutiny their strategies for long-time-horizon experimentation. Another possibility would be to introduce, especially among the largest foundations, a voluntary peer review in which the philanthropic strategies and investments were subject to periodic evaluation by expert peers, be they other foundation leaders or the beneficiaries of grants. Peer review could in principle foster norms that, without the need for formal legal regulation, would help to hold private foundations to a discovery mode. I have in mind the norms that have arisen in the world of academia, where professors with tenure enjoy an unaccountability for their scholarly productivity that is in many respects quite similar to the unaccountability of the assets in a private foundation. Tenure may help to guide scholars toward longer-time-horizon projects than they undertook when they were untenured, and the practice of peer-reviewed scholarship helps to sort better from worse research and creates a forum for reputational competition. Perhaps something similar in the world of philanthropic foundations would be salutary.

The Trailing Edge of Change

Until we pass such reforms, we must face philanthropy as it is, not as it ought to be. How well do actual foundations perform in the United States and elsewhere when measured against the vision articulated and defended here? Are foundations fulfilling their discovery role?

A rigorous assessment is beyond the scope of my argument, but it is worth noting that skepticism is certainly warranted. Many prominent foundation observers, including those who are friends of foundations, believe that they are underperforming when measured on almost any yardstick of success. And they are certainly underperforming if measured by the standards of pluralism and discovery. In 1949, a prominent foundation leader, Edwin Embree, wrote an article for Harper’s Magazine called “Timid Billions,” concluding that, despite obvious social problems and ample philanthropic assets, there was “an ominous absence of that social pioneering that is the essential business of foundations.” More recently, Gara LaMarche, who spent more than 15 years at two of the world’s largest foundations (the Atlantic Philanthropies and the Open Society Institute), concluded that foundations tend to be risk-averse, rather than risk-taking. “Courageous risk-taking is not what most people associate with foundations,” he writes in Boston Review, “whose boards and senior leadership are often dominated by establishment types. If tax preference is meant primarily to encourage boldness, it doesn’t seem to be working.” Joel Fleishman, the former director of the Atlantic Philanthropies and author of The Foundation: A Great American Secret, thinks that foundations would do their work better if they were more transparent and risk-taking. Others, such as Waldemar Nielsen, a prominent author on the subject of philanthropy, have challenged foundations’ support for innovation, arguing that they are more frequently on the “trailing edge, not the cutting edge, of change.” A more recent review of foundation activity suggests that only a small fraction of grantmaking should count as investing boldly in social change.

Perhaps these critics are correct. If so, then so much the worse for foundations, and so much the worse for the distinctive institutional privileges that currently attach to them. My aim here is not to defend the existing behavior and performance of foundations but to identify the right standard by which to assess them—a standard rooted in how foundations can serve democratic societies despite their ineliminable plutocratic aspects.