Book Search:  


Google full text of our books:


Good and Plenty:
The Creative Successes of American Arts Funding
Tyler Cowen

Book Description | Reviews | Table of Contents

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2006, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Follow links for Class Use and other Permissions. For more information, send e-mail to

This file is also available in Adobe Acrobat PDF format

Chapter 1

Warring Perspectives

Many of my conservative and libertarian friends find government funding for the arts unacceptable. They note that after the so-called "Gingrich revolution" of 1994, "we were not even able to get rid of the NEA." They speak of the NEA (National Endowment for the Arts) as the lowest of lows, the one government program that has no justification whatsoever. If such an obvious basket case could survive a conservative Republican Congress, how we can ever hope to rein in government spending?

Most of my arts friends take the contrary political position. They assume that any art lover will favor higher levels of direct government funding. To oversimplify a bit, their basic attitude is that the arts are good, and therefore government funding for the arts is good. They find it difficult to understand how an individual can appreciate the arts without favoring greater public-sector involvement. They lament how American artists are underfunded and undervalued by the state, relative to their western European counterparts.

Why are the two sides to this debate so far apart? How can two groups of people, each well intentioned, look at the same world and see such a different reality?

That people so frequently disagree is a quandary for social science. We might expect that when a person encounters a disagreement with someone, he or she recognizes that the other person, if sufficiently intelligent and honest, is as likely to be right. To paraphrase Garrison Keillor, we cannot all be in the top half of our peer group with regard to wisdom, and presumably we realize this. Furthermore our disputants often have superior training, experience, or raw intelligence. The reality, however, is that convergence of opinion is rare on most political issues. Policy disagreements usually persist, and often deepen, when the individuals engage in sustained personal dialogue. Furthermore the presentation of evidence and the citation of expert opinion usually fail to resolve the dispute.

I write with one foot in the art-lover camp and with another foot in the libertarian economist camp. I try to make each position intelligible, and perhaps even sympathetic (if not convincing), to the other side. I try to show how the other side might believe what it does, and how close the two views might be brought together. Furthermore, I use the fact of persistent disagreement as a kind of datum, as a clue for discovering what the issues are really about.

Except for the 1990s squabbles over the NEA, serious political dialogue on arts policy has simply not taken place. Presidential and congressional candidates prefer not to devote their attention to the issue, except for previous attempts to attack a few controversial grants.

I try to steer the arts policy debate away from its previous focus on the NEA. More significant questions concern the use of our tax system to support nonprofits, creating a favorable climate for philanthropy, the legal treatment of the arts, the arts in the American university, and the evolution of copyright law. I also seek to recast the debate over direct funding of the arts. The central issue is not, as many people suppose, how much money a given governmental agency should receive. It is hard to generate consensus on a question of this kind. Instead a more fruitful inquiry involves what general steps a government can take to promote a wide variety of healthy and diverse funding sources for the arts. For instance should we look more toward direct subsidies or indirect subsidies?

Most of all, arts policy is a window onto how the United States supports creative endeavors. It is commonly believed that we have no arts policy, and on one obvious level this claim is true. No central cabinet-level ministry plans the development of the American arts. At the same time, American governments, at varying levels, have done much to support creative enterprise. The American model arguably mobilizes government more effectively than do many of the European models for arts support. We are further from artistic laissez-faire than is commonly believed to be the case.

I will argue a case for the American system, at least once it is properly understood. The American model encourages artistic creativity, keeps the politicization of art to a minimum, and brings economics and aesthetics into a symbiotic relationship. That being said, the model is not necessarily preferable for all other countries, especially those that trade with the United States and already reap benefits from the American system. Furthermore I do not think it possible to defend each and every aspect of American arts policy. My defense of the American system focuses on the most general features of that regime--capitalist wealth, competitive markets, decentralized and diverse sources of financial support, and indirect subsidies. In the final chapter I also will suggest some means for improving the American system. Insofar as we wish to use direct subsidies, we should restructure them to encourage greater innovation and a greater degree of decentralized support for the arts.1

Finally, arts policy brings into relief critical questions in both policy evaluation and political philosophy. Man is not just a thinking being; he is also an imagining being and a creating being. Cultural policy focuses our attention on how we resolve potential clashes between economic and aesthetic values. Here more than just the arts is on the table. Our ability to reconcile economic and aesthetic values is a general prerequisite of rational policy evaluation.

Moving to yet larger questions, we cannot have a coherent political philosophy without bridging the gap between economic and aesthetic perspectives. For instance critics charge that liberalism cannot satisfy the higher aspirations of the human race. They compare liberal government to an innkeeper who looks after his guests but otherwise has little to offer in the way of vision or a common loyalty. On the international scene, the United States is often seen as a military and economic behemoth, but as lacking in concern for cultural values or beauty. I wish to put this picture to rest, and to reclaim America's rightful role in offering a liberal vision for beauty and creative human achievement.

I will use arts policy to begin a new sketch of a liberal state. The public sector can encourage a proliferation of diverse cultural outputs and in that regard offer a rich menu of life-enhancing options. At the same time, we do not have to abandon the values of free speech and neutrality across (noncoercive) competing lifestyles. All this can be done in a manner consistent with prosperity and other economic objectives. A state--in particular the American state--can be involved with matters aesthetic without losing its liberal character. We also will see that, counterintuitively, a rich diversity of artistic achievement is compatible with the ideas of cultural centrality and the use of culture to bind a polity together.2

Since the potential clash between economic and aesthetic perspectives is our central motivating concern, let us now turn to it more directly. We will look at policy evaluation at the microlevel, and then throughout the course of the book build back toward larger visions.

Aesthetics and Economics

Neither the aesthetic approach nor the economic approach can stand alone as a tool for evaluating policy. Rather than try to elevate one approach at the expense of the other, I will look for policy recommendations that might reasonably persuade a person who assigned some validity to each perspective--economist and aesthete--at the same time.

The economic approach seeks to give consumers what they want. Many people watch The Sopranos, but few come home and pick up James Joyce's Ulysses after a day of hard work or even after a day of easy work. A book does not have to improve on successive rereadings if it will be read only once. Consumers care mostly about fun and convenience. A survey indicated that 46 percent of Americans would be unwilling to give up television for the rest of their lives in return for a million dollars.3

More technically, I define the economic approach in terms of standard microeconomics and Paretian welfare economics. Economists measure value in terms of willingness to pay or willingness to be paid.

The aesthetic approach favors products that will "last the ages." It points our attention to study-intensive classics and canonical works. It is harder to point to a single canonical presentation of the aesthetic approach, but I define the aesthetic view as suggesting that quality culture has intrinsic value. Matthew Arnold wrote that "culture is, or ought to be, the study and pursuit of perfection; and that of perfection as pursued by culture, beauty and intelligence, or in other words, sweetness and light."4

Under the aesthetic approach the notion of a just and beautiful society is prior to the value of satisfying individual preferences. Often a vision of "cultural democracy" is paramount. Art has elevating and developmental powers, and in this view all democratic citizens have a right to such experiences. Government is seen as capable of carving out a realm for human education and freedom, removed from negative commercial pressures. Without irony Hans Haacke wrote: "If culture is to remain free, we taxpayers must be ready to finance it!"5

Under the economic approach, high art is simply another minority taste and holds no special normative status. The economic approach treats the critic as simply another consumer, albeit with different intellectual and material endowments. A particular critic may be a consumer with time on his hands, an academic job, and special training in the classical piano repertoire. For the economist, this point of view is no more privileged than that of the consumer who works two jobs, has three small children, and never finished high school. Willingness to pay, rather than social standing, education, or any other variable, determines how much a person's preference is to be weighted. Along these lines, one study found that the American public had more respect for bus drivers and baseball players than for art critics (or poets, ballet dancers, or professional actors, for that matter).6

Often critics include popular culture in their canons. But they still judge popular culture from the perspective of a critic, rather than from the perspective of a consumer. These critics are trying to invest popular culture with the same qualities they have found in high culture. The critical commentary on Seinfeld, for instance, does not pursue the form that fan discussions of the show would take, such as who is the funniest character, whether Jerry and Elaine should reestablish a romantic relationship, or whether Kramer will get a job. It is possible to publish an academic article on "the postmodern in Seinfeld," but not on whether Elaine should have split up with her last boyfriend, although the latter is of greater concern to most of the show's viewers.

By siding with the consumer point of view, economics serves as a radical attempt to demystify the perspective of the critic. That is one reason why critics do not like the economic approach.

The Problem of Commensurability

The aesthetic approach has a hard time making aesthetic values easily commensurable. It is difficult to decide whether Shakespeare's Hamlet is better than his King Lear, and even harder to persuade others of our decision or define what such a ranking would mean. How many Gershwin songs sum up to a Shostakovich symphony? Is a Haydn string quartet better than a Hemingway short story? How does a Blake poem compare to a modern ballet performance? When evaluating a government policy or a cultural era, do we look at the peaks that result, the total amount of art, or some weighted average of the two?

The average quality of output provides an inadequate standard of evaluation. The majority of artistic projects are commercial and aesthetic failures, regardless of the source of funds, the political system, or the time period. We care more about the successes than about the fate of every cultural undertaking. Similarly, we cannot judge government funding, or laissez-faire for that matter, by looking at the median or typical project. Anecdotes of failure can easily mislead. We should not be surprised if a particular system produces much junk and offensive material.

The most common benchmark uses "peaks" to compare one culture to another. Government funding is praised, for instance, for having supported Bach, Velázquez, and Edmund Spenser. The same invocation of peaks has been used to compare "the moderns" to "the ancients." We might ask what modern composer compares to Beethoven or what modern poem measures up to Homer's Odyssey. Or we might ask, "Which age has produced the best symphony?"

Yet the peaks standard demonstrates a potential bias against the market and prejudges significant, and open, aesthetic questions. Why should the greatness of the best composer, or the best poet, be the relevant unit for judging a culture? What if one culture (modernity?) produces lesser creative titans but produces many more of them? How are we to weigh the quality of the peak versus quantity of the total?

It is also an open question what is the right unit for judging a peak. Instead of looking at the highest peaks, we could judge an era by how good its "one hundred best composers" are, or by the aesthetic worth of its "best five thousand hours of music." Or consider a peak of a different kind: "How many excellent musical genres does an age have?" By these standards, contemporary times fare better, vis-à-vis the era of Beethoven, than if we just compare the best composer from each period. We have many talented composers today, in many different musical fields, even though today's best composer is not the equal of Beethoven.

Why the focus on a single artistic work and its greatness? Mozart's Don Giovanni has musical beauty, terror, comedy, and a sense of the sublime, making it a favorite of opera connoisseurs. But what if consumers draw their comedy from one work, their terror from another, their beautiful music from yet another, and so on? Artistic peaks typically bundle qualities together. Yet arguably a world with unbundled qualities is superior, since it allows consumers to pick and choose how much of each quality they want, and from which source.

We cite "peaks" when making an aesthetic assessment because they are relatively easy to observe and talk about. Few individuals know much about eighteenth-century culture except for its peaks. But the peaks standard remains incomplete. The notion of a peak does not correspond to how much aesthetic value is produced in an era or to how much that value is enjoyed.

The economic approach, whatever its weaknesses, has one major virtue. It makes aesthetic values commensurable, at least in principle. Individuals' willingness to pay gives us a standard, however crude, for how many Gershwin songs are worth a Haydn string quartet. So if individuals, taken collectively, are willing to pay $100,000 for a new Haydn recording, and $200,000 for a new Gershwin recording, the economic standard judges the Gershwin disc to be worth more. The economic approach rejects the dominance of aesthetic peaks per se. It looks instead at how much total value is produced, again as measured in monetary terms. Money is not all that counts, but all things that count are measured in terms of money.

The economic method therefore offers a practicable response to the awkward questions of how to define peaks, or how to weight quality versus quantity. This relatively straightforward approach to commensurability, no matter how much it horrifies the art purist, means the economic approach will be invoked repeatedly to adjudicate disputes. Analogously, philosopher and economist Henry Sidgwick argued that utilitarianism, whatever its apparent failings, was the only philosophy that could sensibly weigh competing values at the margin. Economics, a modern form of utilitarianism, offers a similar property. Perhaps the economic approach will be battered with telling and persuasive critiques. Yet it will reemerge in some form to address commensurability, even if it does not reign as the sole standard for judging policy.

The economic approach also holds a core of common sense at its center. Art lovers sometimes write or talk as if economic costs do not matter. They tend to evaluate regimes in terms of the quality of the art that is produced, without considering the opportunity costs of that art. More and better art is equated with a better society. We are never told how many bags of potato chips, or how many antipoverty programs, we should sacrifice to receive another great artistic performance, or how we might hope to find out such an answer. The economic approach reflects the view of the common man that art is not everything, or even the most important thing.

On the negative side, the economic approach considers only a limited range of values, namely those embodied in individual preferences and expressed in terms of willingness to pay. This postulate is self-evident to many economists, but it fails to command wider assent. It wishes to erect "satisfying a preference" as an independent ethical value but is unwilling to consider any possible competing values, apart from preferences. It is hard to see why nonpreference values should not be admitted to a broader decision calculus.

Typically economists retreat to their intuition that satisfying preferences is somehow "real," and that pursuing nonpreference values is religious, mystical, or paternalistic. The rest of the world, however, has not found this distinction persuasive. They do not see why satisfying preferences should be a value of special and sole importance, especially when those same preferences may be ill informed, inconsistent, malicious, or spiteful. The decisions to count all preferences, to use money as the measuring rod, and to weight all market demands equally must themselves rely on external ethical judgments. For that reason, the economist has no a priori means of dismissing nonpreference values from the overall policy evaluation.

Economists often accuse other methods of policy evaluation of paternalism, but this is more of an ad hominem attack than an argument. It is paternalism to force an alternative on society that nobody wants, or to force an individual to do something against her will. It is less clear why it is paternalistic to use nonpreference-based values as "tiebreakers." What if social preferences do not themselves suggest a clear or nearly unanimous rank ordering? Why is it so unreasonable to look to other values to resolve deadlocked preferences?

Economists have long treated merit goods as a kind of "add-on" to standard cost-benefit analysis. That is, we stick with standard cost-benefit methods and simply add on a value for the relevant merit good. By definition, a merit good is of greater value--for intrinsic reasons--than is reflected in individual preferences. This procedure, however, requires a common understanding on the relevant value trade-offs. It also requires that noneconomic values be squeezed into the economic framework. Many art lovers wish to keep the idea of the merit good but jettison the underlying economics.7

Some economists prefer contractarian ideas, to derive unanimity at a broader level of social choice. Yet we must specify what information and values people will have "behind the veil," thereby also determing what they would choose. We first need some prior theory of what is right, or some prior theory of what kind of life a self-interested person would pursue. But if we have significant prior knowledge about these questions, we should start with it directly. Such foundations, if present, would move us away from the economic approach.

The economic approach also performs best when the distribution of wealth is given, preferences are fixed and well defined, and the policy change in question is a small one. Perhaps economics can tell us whether we should build an extra stadium in St. Louis, but it is less well suited for analyzing the entirety of arts policy. Often art consumers do not know what they want until it is before them, which vitiates the fixed-preference assumption. The arts and thus arts policy are shaping what preferences will be. More generally, the supply of art helps determine what "languages" individuals will have at their command for interpreting social reality. The economic method, taken alone, is poorly suited for evaluating such choices.8

Pure Paretian improvements, which make everybody better off, are hard to come by. So for most practical purposes, applied welfare economics boils down to wealth maximization. But then we must wonder why wealth is such a supreme value. Even Richard Posner gave up his wealth maximization standard, in response to scathing criticism from Ronald Dworkin. Dworkin made the simple point that wealth cannot plausibly be an end in itself. Wealth may be a useful proxy for other values, such as justice, fairness, happiness, and beauty, but if so (and these connections are not self-evident), we should target these other values directly. These targets may be hard to measure, but if they were impossible to measure, we could not claim that wealth was a useful proxy for them. We would again have no rationale for the wealth maximization standard.9

Arthur Schlesinger, Jr., remarked that only "a moment's reflection" is needed to realize that the economic approach to arts policy is "absurd." While this is a gross overstatement, the economic approach has not won many debates in the philosophic arena.10

How Do the Differing Approaches View Subsidies?

The aesthetic approach opens the door for a case for government subsidies to art. Few critics believe that market-driven culture will maximize its potential aesthetic value, or admit of no aesthetic improvements. The perspective of the critic therefore finds an obvious, and potentially remediable, flaw in market outcomes. To the extent that the government can use good taste to target outcomes, direct subsidies could improve on laissez-faire.

The economic framework, by adopting the perspective of consumers, leads to greater skepticism about subsidies. Consumers spend most of their cultural time with commercially viable products. They watch television, buy popular novels, and listen to popular music; esoteric forms of high culture are of less interest to them. Not only will art subsidies remove resources from nonartistic pursuits, but subsidies may cause the quality of art to decline, from the perspective of consumers. John Updike wrote: "Government money in the arts, I fear, can only deflect artists from their responsibility to find an authentic market for their products, an actual audience for their performance."11

The economic approach stresses the notion of opportunity cost. Subsidies are a good idea only if, once all effects are taken into account, their dollar-valued benefits outweigh their dollar-valued costs.

The Federal Writers' Project of the New Deal supported Saul Bellow, Richard Wright, Ralph Ellison, and Zora Neale Hurston, among other individuals who later became noted writers (see chapter 3). Many literature lovers then conclude that the program was a good one. The economist, in contrast, could point out that the Federal Writers' Project spent $27 million for about one thousand books and pamphlets, at a cost of $27,000 per publication. Translated into dollars for the year 2000, this amounts to $337,500 per publication, or $337.5 million for the total. While some of the publications were of very high quality, they cost a great deal. It is hard to believe that offering indiscriminate literary advances of $337,500 per work is a worthwhile investment, even if it produces some first-rate books. Frederic Bastiat, writing in the nineteenth century, argued that proponents of arts subsidies typically focus on "what is seen" and neglect "what is not seen," namely the other projects and outputs that would have been funded with the money.12

One estimate (Wyszomirski 1999, p. 134), aimed at defending the New Deal's Works Progress Administration (WPA), suggested that the $5 million invested in the five core art projects of the WPA produced artistic outputs later worth $450 million. In reality those investments did not make overwhelming economic sense. Had the $5 million been invested in the stock market at the end of 1934, as of December 1999 (1999 is the date of publication of Wyszomirski's book), it would have yielded more than $773 million. Of course the comparison is highly sensitive to the chosen starting and ending dates, and if we picked other dates, the investment in paintings might look better. Furthermore these pictures yielded enjoyment in the interim. Nonetheless the most acclaimed arts program in U.S. history--the WPA--was not an overwhelming success in economic terms, if we compare it to alternative investments.13

In the economic framework, the arts should be subsidized only when they produce significant "positive externalities" at the relevant margin. A positive externality occurs to the extent that artistic production benefits individuals other than the paying customers. For instance, the arts may improve the social and moral climate of the nation or may spur more general forms of economic development. When such externalities are present, the total social value of art is greater than what people are willing to pay for in the private marketplace. A subsidy, if applied correctly, can induce more socially beneficial art and create more resources than it costs.14

Platitudes about the wonders of the arts do not constitute an argument for an externality at the relevant margin. The arts in general, by bringing "sweetness and light" into our lives, make the world a much better place, above and beyond what we have paid for them. This point, however, does not suffice to drive policy. The question is whether more art, at the margin, will do much to improve human welfare. If an arts council funds an extra show at the Whitney Museum, the rate of economic growth in New York is unlikely to increase.

In the absence of such a marginal externality, the economic approach implies that artistic production is best left to the market. Consumers will then buy just enough art until the social value from producing more art, as defined in terms of willingness to pay, would equal the social cost of producing that art. No further improvement in human welfare would be possible.

In contrast, most subsidy advocates start by asking whether art subsidies make our society morebeautiful. We then can weigh whether this beauty is worth the cost in terms of efficiency, but efficiency alone does not prejudge the matter. In this view art has a value above and beyond what people are willing to pay for it. It may be better, all things considered, to have a less efficient but more beautiful society.

Many subsidy advocates go further. They conceive of policy benefits, including health, safety, and prosperity, primarily in aesthetic rather than in economic terms. They are evaluating all of society, not just arts policy, from what is primarily an aesthetic point of view. A wealthy society is presumably a precondition for a healthy society, and in that regard efficiency plays a role in supporting the aesthetic. For that reason, the typical subsidy advocate does not (or at least need not) ignore efficiency. But many subsidy advocates subordinate efficiency to the aesthetic, whereas the economist subordinates the aesthetic to the individual preferences behind the efficiency standard.


The best-case scenario would allow for a rapprochement between the economic and aesthetic perspectives. If we examine the evidence and positive arguments with sufficient care, they might produce agreement on a number of practical questions. The remaining disagreements might be smaller and more manageable than the broader philosophic issues raised above. Through this method we might make progress on the concrete questions of arts policy.

The path here is a bit complex, and takes the following outline. First I start with some purely economic arguments for (some kinds of ) subsidies, such as the economic development argument and the free-lunch argument. I will downgrade their importance, and they will not reemerge as factors to shape policy. Later parts of this chapter proceed to two issues--the decentralization argument and the prestige argument--that have force in both economic and aesthetic terms. So when using these arguments to shape policy, I hope to show that the relevant conclusions will be supported by a confluence of economic and aesthetic perspectives. We can therefore sidestep some of the stickier philosophic issues.

Once these arguments are laid out, the three following chapters will examine which institutions do the best job in light of the decentralization and prestige arguments. These chapters will cover the practical topics of direct subsidies, indirect subsidies, and copyright protection for the arts. The final chapter will tie together the philosophic and empirical investigations with some concrete policy recommendations, some remarks about the liberal state, and an account of why agreement is so hard to come by in this area.

The Economic Development Argument

Many externalities arguments fail to establish a strong case for government subsidies. For instance subsidy advocates frequently cite the economic benefits of a healthy artistic community for a city. The arts generate employment, attract tourists, and produce tax revenue. All these effects may contribute to the general social welfare. The English composer Henry Purcell was an early proponent of arguments of this kind. In the preface to his opera The Fairy Queen, he claimed that opera subsidies would draw tourists and make London an economically thriving city.15

Insofar as the arts spur economic development, the decentralization argument, discussed further below, captures the relevant truths, at least in the American context. To anticipate that claim, the most likely positive scenario is when a private or political entrepreneur spots and invests in an undervalued artistic cluster. For this result we must look toward multiple and varied sources of financial support. No single firm, agency, or individual has proven adept at picking artistic winners across a large span of human achievement. More generally, the production of new and diverse ideas--cultural and otherwise--will contribute to economic growth. Given that we are keeping the decentralization argument on the table, we can set aside economic development as a separate concern.

More generally, the economic benefits of the arts do not themselves necessarily translate into an argument for subsidy. First, the ability of the arts to attract tourists, generate employment, and produce tax revenue implies that those arts are likely to be economically viable in their own terms. Furthermore the arts add value to a region, but a region adds value to the arts. We therefore can expect cities and the arts to come together, even in lieu of a subsidy. In the language of economics, getting the arts into cities might be a game of coordination or a question of finding a common locale. Most urban economic activity both produces and receives synergistic benefits (thus the willingness to pay higher city rents), but this does not imply that all urban activities require government subsidy.

Second, social benefits are not unique to the arts. Most productive activities bring social benefits to large numbers of people. Most kinds of economic activity create new jobs, raise the tax base, and contribute to general prosperity and social order, all of which yield positive externalities. The relevant policy question is, not whether the arts involve some positive externality, but whether the "economic development externality" from the arts is greater than from alternative investments. In fact, if the arts require subsidy to flourish, they are unlikely to be an especially strong engine of economic growth, whatever their other virtues.

Some activities have to be the ones that provide the greatest economic stimulus to localities, and of course this may include the arts. Nonetheless we have no firm evidence that the arts, on average, rank especially high on this list. Even if the arts stand well above the average, it is our duty, as policy analysts, to promote the best alternative on the menu. In contrast, the emphasis on decentralization and entrepreneurship looks for instances of undervalued resources. The arts in general need not be undervalued, but good institutions can still help us find particular instances of undervalued artistic creations.

We should be skeptical of "economic impact" studies that show the importance of the arts to a community. A study of this kind might show that an arts festival or new arts arena brings millions of dollars in economic value. But these studies typically treat arts expenditures as creating value out of nothing. Implicitly it is assumed that if the money had not been spent on the arts, no other economic or social values would have been produced. Again, the relevant comparison is whether an arts arena leads to more value than some alternative. When we look at economic impact studies for one industry at a time, they all appear to show high benefits. But this means that the net benefits of any single project are low, zero, or perhaps even negative on average. By investing in one good idea we are always forsaking another good idea. In essence those studies list gross benefits rather than net benefits. Furthermore, once an economic impact study is being done, the resources are likely no longer undervalued.16

To compound the inaccuracies, economic impact studies assume a "multiplier" effect. Under the multiplier hypothesis, the expenditure of a single dollar leads to a creation of several dollars' worth of value. To provide an example, the construction of an arts arena helps nearby restaurants, which in turn increases the demand for waiters, which in turn sends more money into the community, and so on. As the argument goes, the expenditure of a single dollar on the arts arena might create three or four dollars of value, once the money filters through the economy. Even if we accept the logic behind the multiplier effect, it does not demonstrate a relevant externality. Had the arts arena not been produced, the money would have stimulated some other set of demands and created some other set of multiplier effects. The net multiplier effect from the arts arena still might be zero or negative, unless it can be demonstrated that the arts are especially productive. No such demonstration, however, has been offered.17

In addition a community must invest resources to attract artistic activities from other regions. If the arts do bring net benefits, these benefits will, to some extent, be reflected in the cost of bidding for those activities. To provide a simple analogy, Silicon Valley is an economic power-house. Nonetheless the costs of bidding Silicon Valley into Connecticut would probably exceed the benefits.18 Furthermore when cities court sports teams, arenas, and arts donations, often they are bidding those resources away from other cities. In this regard the gains of a single city (or country) do not always bring net gains for the world as a whole.

We do see particular cases where artistic heritage has an especially high and obvious economic value, and centralized action is required. Consider the city of Venice. It is hard to believe that Venice would have a healthier economy if it paved over its canals and turned that space over to street vendors. Venice therefore has no comparably effective way of investing the money it spends on preservation. So this is a good argument for subsidizing the historic preservation of Venice, but it does not provide a good reason to subsidize the arts more generally.

The "Free-Lunch" Argument: Small Sums vs. Large Sums

The NEA, like many other governmental subsidies, looks best when we measure its cost in per capita terms. NEA expenditures, at their height, were no more than 65¢ for each American citizen. Currently the figure is below 50¢ a person, and it is unlikely to rise dramatically in the foreseeable future.

Some subsidy advocates talk as if subsidy programs literally cost us nothing or only a negligible sum. The economist, brought up to believe there is no free lunch, typically considers this argument absurd. Nonetheless it reemerges frequently in the intuitions of art lovers, so let us look at it more carefully, distinguishing between more and less extreme forms of the argument.

Under the more extreme argument, the taxes for the NEA literally cost us nothing, once we translate them into per capita terms. While individuals have less money to spend, the sum forgone is small. Arguably there are no major individual "life projects" that will fail to be achieved, for want of the 50¢ or even for want of a few dollars (keep in mind that most poor Americans do not pay income taxes at all). Most people have some amount of financial slack built into their plans. They die before spending all their money and do not care whether their bequest to their children is a few dollars more or less. Perhaps the small sums are never noticed one way or the other. Even in Germany, which has among the highest direct subsidies per capita of the arts, the sum falls in the range of $80 per person per year.

In contrast, government subsidies bring large and visible benefits in the form of discrete projects. The beneficiaries notice the difference, as the world receives some artworks that otherwise would not have been produced. These artworks benefit relatively small numbers of people, including the artist, but the benefits are significant for each recipient. Most of the benefits lie well above the threshold of noticeability, and thus they increase human happiness.

The less extreme form of the argument, rather than finding a free lunch, relies on antiegalitarian intuitions. We take small amounts from many individuals to give concentrated benefits to a few, and indeed often to the relatively wealthy. The distribution of goodness in society becomes less equal, but our peaks of aesthetic achievement become more beautiful. Subsidy supporters do not like to publicize this reality, but these antiegalitarian intuitions are central to many of the arguments for subsidies.

In more academic terms, the philosophical doctrine known as "perfectionism" suggests that societies should seek to reach their highest possible peaks, even if this requires a sacrifice of equality. The threshold argument and the perfectionist argument operate together to reinforce intuitions in favor of subsidies.19

If we accept this portrait, subsidies appear to be a good deal from the standpoint of an outside observer. In terms of real human happiness they cost either nothing or close to nothing. At the same time they create concentrated pockets of joy and aesthetic wonder. Many of the costs fall below the noticeability threshold, and many of the benefits lie above it. It is no wonder that the logic of subsidy meets with relatively little resistance from taxpayers when applied on a case-by-case basis.

We could, of course, justify many subsidies on these grounds, not just subsidies to the arts. We could tax each person a penny and redistribute the funds to a few hundred or to a few thousand people. No one would bear much of a real cost in terms of happiness, but the fund recipients could undertake new and valuable projects. A sufficiently small tax, applied to a broad enough base, will involve very small per capita costs.

The threshold and perfectionist arguments for subsidies have persuasive force, but they also prove too much. As stated above, they could justify virtually any small subsidy, provided that subsidy encouraged new projects of discrete value.

Subsidies appear less attractive when evaluated as a bundle. The taxes I pay for small subsidy programs do not make me worse off, even putting aside the artistic benefits I reap from the subsidies. But I am hurt by the taxation of 35 percent of my income. Much of this 35 percent is composed of very small subsidies to particular individuals, corporations, and institutions. Any one of these subsidies may make sense when evaluated individually, but they involve a real cost when evaluated as a bundle. As a bundle, they substitute governmentally determined projects for the projects I would have chosen.

Subsidy supporters usually prefer to present and evaluate the subsidy in stand-alone terms. Subsidy critics portray arts subsidies as part of a larger pattern of largesse and unnecessary redistribution. They see subsidies as part of a general trend to substitute political allocation for voluntary individual allocation of resources.

Similarly, subsidies appear less attractive if they are evaluated in terms of concentrated costs, not diffuse costs. For instance, it may cost, at most, a few hundred dollars of investment to protect a child, in the underdeveloped world, against a fatal or debilitating disease. For the sake of argument, let us round this number up considerably, to $2,000.20 Under this reckoning, if we consider reallocating NEA funds to this end, NEA arts programs cost over fifty thousand lives each year. We could, in principle, abolish the NEA and distribute the funds to Haiti or India, using a private charity if we do not trust traditional mechanisms of foreign aid. Or we could abolish the NEA and every year give more than ten thousand poor Americans $10,000 each, to spend as they wish. In a later chapter, I shall refer to this as the "What about the Haitians?" critique of arts funding, and it is a challenge that any prosubsidy view must face.

Subsidy advocates, when they compare arts expenditures to discrete alternatives, frequently have chosen examples from the military (NB: this attempt to slant the comparison has become less fashionable since 9/11). The NEA, for instance, costs us only a fraction of an aircraft carrier or a nuclear submarine. Even a simple M-1 tank costs over $3 million, which means that fewer than forty tanks would equal the NEA budget.

Which is the correct comparison to make? Do NEA subsidies cost us each less than 50¢? Do they cost thirty-five or so M-1 tanks? Or do arts subsidies cost the lives of thousands of innocent children? And should we evaluate the subsidies in stand-alone terms or as part of a larger bundle of programs?

The relevant policy alternative will depend on the political climate of the day. But we should not proceed by comparing arts subsidies to the least attractive opportunity for those funds, such as the possibility of "an extra 50¢ for everybody." To the extent we can "play God" with the government budget--the very presupposition of policy analysis--we can spend those dollars as we wish. If we accept the threshold and perfectionist arguments in the first place, we would have better options than simply leaving 50¢ in the pocket of each individual.

We started by presenting the free-lunch argument as a potential reason for arts subsidies. But it is also a potential reason not to have such subsidies. Under one interpretation of the choices before us, we are condemning thousands of innocent children to death and disease when we subsidize the arts.

I now consider two other arguments for government involvement in the arts, arguments that I consider to be more significant--the decentralization argument and the prestige argument.

The Decentralization Argument

In some areas of human life, we learn by amassing the cooperation of "the best and the brightest" through centralized institutions. The Manhattan Project proceeded this way in its later stages. Or if we wish to study quarks, it may be best to invest heavily in a single, high-powered particle smasher (though not all scientists agree on the cost-effectiveness of this approach).

Other endeavors require more decentralization. Artistic discovery, for instance, is rarely a matter of brute force, or amassing enough laborers to work on perfecting a single technology. Rather we are hoping that the artist can "look at things differently" and see something that others have not. To make this happen, the artist must have the ability to market his or her vision to a diverse set of consumers, donors, and funders. It is unlikely that any single source of support will grasp the importance of all these innovations. Creativity flourishes when many different visions have a chance of succeeding.

Along these lines, we can view arts funding as a portfolio or investment problem. In most cultural markets, if we are trying to pick tomorrow's winners, we cannot forecast in advance what will work. In this regard cultural markets resemble Internet start-up firms or classic R & D problems. A few tries will hit it big, and many more will fail. In this kind of environment it makes sense to try many different approaches, rather than put all our eggs in one basket.

The American system helps generate artistic innovations, encourages new ways of marketing and distribution, and supports competing critical visions for artistic contributions. In essence, the American system satisfies the "Hayekian" standard that institutions should support the generation and dissemination of knowledge. Austrian economist Friedrich A. Hayek emphasized "competition as a discovery procedure" in many of his writings, and stressed the inability of a central authority to plan discovery. The market has the virtue of mobilizing decentralized knowledge. Entrepreneurs have opportunities to test their diverse visions in a setting with many differing sources of financial support.

Hayek's argument has often been viewed as a plea for laissez-faire, but a look at arts policy belies the necessity of that interpretation. In reality, the argument implies that we should have many decentralized sources for producing and evaluating ideas. This may or may not imply laissez-faire, depending on the institutional setting. Both tax breaks for knowledge-producing institutions and a publicly subsidized university system may encourage decentralization, to provide two examples. American arts policy uses government to induce a more decentralized pattern of financial support than would arise through pure laissez-faire.21

These policies do not imply that investments in the arts, relative to alternatives, yield especially high social returns. We should not think in terms of subsidizing the arts at the expense of other activities, or giving the arts special status. Rather we should think of American policy as encouraging decentralization for all creative activities, the arts included.

In this regard the development and decentralization arguments are distinct. The development argument asserts that the arts bring net economic advantage at the relevant margin. The decentralization approach seeks the greatest possible chance of generating, at the relevant margin, whatever brings the greatest net economic advantage.

Most deliberate governmental attempts to stimulate the discovery process have failed, and for reasons that Hayek and other economists have outlined. Government does not have the knowledge needed to centrally plan innovation. To provide one well-known example, after the energy crisis of the 1970s the U.S. government subsidized research into alternative energy sources, such as synfuels and solar energy. The end result was wasted money and little or no net technological progress with energy conservation. The government had no idea which energy-saving technologies were going to be the winners. Most improvements in energy efficiency have come from market-based institutions, encouraged by the desire to save money or to earn a profit from a new technology.

Governments usually stimulate discovery best when they eschew central planning, instead providing support according to some non-market criteria. This approach does not require that government can do an especially good job of picking winners, or that government is smarter than the market. It requires only that government distribute its support according to some principle differing from what is already available. The real question is not whether decentralization is beneficial in today's world, but rather how we should encourage decentralization at the margin.

Why We Should Support Decentralization

At least three arguments suggest that government should, either directly or indirectly, help markets achieve a more effective degree of decentralization.

The first argument for decentralization invokes the aesthetic approach. In this view more art is desirable, especially if the new art meets high critical standards and can stand the test of time. We therefore should invest in the preconditions of quality art, namely diverse sources of financial support.

The second argument for decentralization invokes the perspective of the economist, and Paretian welfare economics in particular. Information is to some extent a public good and involves a positive externality. That is, creators do not reap the full reward of their labors.

In the nonartistic realm, the discoverers of quantum mechanics created great value but were paid relatively little. Thomas Edison was one of the wealthiest inventors in his time, but he reaped only a small fraction of the value he created. In the arts, the great creators make possible whole artistic genres and revolutions, yet they do not receive compensation for these indirect effects. Shakespeare, Mozart, and Beethoven all made a good living in their lifetime, but they did not reap anything close to the full value of their labors. Each offered lasting contributions to other artists and to the human race. Picasso was much wealthier than these creators, but still he received money only for his own paintings. He received nothing for enabling the innovations of Braque, Gris, and many other modern artists.

The greatest creators are undercompensated relative to their efforts. So if a mechanism encourages first-rate creation, it will find support from an economic perspective as well. Information is a public good with positive external benefits. Once it is produced, it typically spreads more widely than the initial creators can charge for. One estimate (Nordhaus 2004) suggests that creators receive no more than 5 percent of the value of their innovations.

Copyright does not eliminate this imperfection, even when it can be enforced. Copyright law is based on the distinction between an idea and the expression of an idea. At most it protects rights to the expression, but not to the idea itself. So copyright law can stop someone from reproducing an image of a painting, but it gives the creator nothing when others learn from his or her inspiration. Ideas remain undersupplied, even when the available artistic products make their way in the marketplace. That being said, an effective copyright regime is a good spur to decentralized creativity.

The third argument for decentralization also comes from economics. Most extant analyses suggest that entrepreneurs undersupply variety to consumers. Assume that an entrepreneur is thinking of marketing a new kind of music, say in the form of a compact disc. Buyers will value that disc from, say, a range of $2 to $40. The seller usually cannot price-discriminate perfectly and thus will pick some price in the middle, say $15. Many consumers are thus reaping more value than they must pay for. Entrepreneurs will not take these extra valuations into account when deciding to introduce new products, as they care only about what they can charge people for. Too few new products will be introduced, again by the standards of Paretian welfare economics.22

Note that the weight of these arguments does not suggest subsidies to art per se. Perhaps we have too few great creators, and too little variety. But we cannot make any direct leaps from these claims to the topic of art. Directly subsidizing more variety in art, for instance, would mean less variety in other areas. We might have less variety in foods, cars, pencils, or scientific discoveries. No argument has been made that more variety in art, or ideas in art, is more important than, say, more variety in science. For that reason, arguments two and three do not show that we should target art in any particular way. At most the arguments establish that we should encourage decentralized financial support for all creative activities, the arts included. This point will prove critical as the argument proceeds.

We see also that the two perspectives behind the decentralization argument--economic and aesthetic--offer slightly different rationales for subsidy. The critic's perspective says that decentralization has produced good art, so we should support an effective art-generating mechanism. It does not ask whether we are investing too much in decentralized support for creativity, relative to alternative uses for those resources. The economist's perspective implies we should invest more in creative discovery, relative to what a pure market would bring. It does not refer directly to art, though of course the notion of creative discovery includes art.

The Prestige Argument

One version of the externalities argument cites national, regional, or cultural prestige. Many citizens take pride in the fact that their country produces art of a certain kind, even though they are not willing to pay for that art as customers. They like the idea that their society excels in creative activities, at least provided those creations are prestigious ones. Martin Archer Shee, writing in 1809, argued that governments should support the arts, because societies win their historical prestige through their artistic contributions; he referred to "living dignity and deathless fame."23

In similar fashion, many British and French citizens supported public subsidies to the Concorde in Britain and France, even though they never expected to fly in the plane. We salute flags, sing national anthems, and sometimes even fight wars, all in the name of national symbolism and prestige.

In similar fashion, we favor arts policies partly for symbolic reasons. Subsidy advocates want an art that has a link to the public sphere. They like the idea that their government has a hand in something noble. They want to be proud, not only of their country and their arts, but also of their governing body. A government that supports the arts is seen as more beautiful and more prestigious. The supporters of that state can feel good about themselves as voters and citizens. The French in particular appear to hold this motive. They will on one hand purchase American popular culture but in the political arena support trade restrictions on those same products and heavy subsidies for French art.

For this reason, it is often beside the point to argue that the arts will flourish without direct subsidies in the instrumental sense. Even if the claim is true, it does not satisfy the strong preference of many to have a state that honors creativity and backs up that honor with direct expenditures.

Furthermore many observers find something objectionable about the idea of allowing the arts to fall exclusively within the private sector. They feel that the arts should, to some degree, be brought within the push and pull of democratic discourse. These individuals want their government to make a statement that the arts are about more than simply money and about more than mere contractual agreements. They want the arts to be viewed as a vital part of a national or regional heritage, and one that receives official recognition from the government as such. Recognition of this kind requires government financial support and government arts programs. Walter Benjamin's (1986) famous essay "The Work of Art in an Age of Mechanical Reproduction" spoke of artworks as surrounded by an aura, but this phenomenon extends to the public sector as well. Government subsidies to art help produce the aura of democratic politics, especially in western Europe.

To cite another example of this principle, national parks also have an aura. Many people think it wrong that a private individual should have the control and dominion over nature that private property implies. They favor national parks, not only because they want the parks to be freely available, but also because they want the parks to be publicly owned. Publicly owned parks constitute a statement that the bounty of nature is, as a matter of right, freely available to all. In this view, the case for public ownership is not vitiated by the fact that the parks are patronized by the wealthy more than by the very poor. Rather it is simply unacceptable for one individual to invoke a private property right to exclude another individual from a more generally owned natural heritage. Government ownership of parks follows from this attitude.

Note that the prestige argument can be given an aesthetic or an economic interpretation. In the language of aesthetics, such a state and polity is nobler and more beautiful, and thus desirable for its own sake. In the language of economics and externalities, the argument for government art subsidies is that many individuals want government support for the arts. By subsidizing the arts, the government satisfies these urges for a particular kind of society and a particular kind of government. This kind of satisfaction, by definition, cannot be procured through private market transactions. The marketplace cannot produce "government support for the arts," no matter how well market institutions might work in the traditional economic sense.

The prestige argument also relies on a different implicit notion of subsidy than does the decentralization argument. For the decentralization argument to work, the subsidy must actually encourage the extent and diversity of creative enterprise. Many subsidies, of course, do not, as we will see below. For the prestige argument to work, the subsidy need only be seen as a subsidy in the eyes of those enjoying the prestige. It should be as visible and as public as possible. The subsidy need not do much to encourage artistic creation, provided it appears sufficiently noble (of course to some extent image and status follow from instrumental success).

We also see why subsidy advocates describe NEA expenditures in terms of per capita costs--so many cents per head--rather than comparing the money spent on the arts to the number of poor Haitian children who could have been saved. A government is more prestigious if it funds the arts from expenditures on beer and potato chips. A government that appears to fund the arts "on the backs" of starving Haitian peasants appears less noble and thus delivers an inferior result to its supporters. In this framework the value of the program is closely related to how we describe the program. To the extent that arts funding is about producing prestige, its advocates will attempt to describe it in prestigious terms. They do this not merely to lobby or to persuade others; rather the favorable description is itself part of the desired result, namely a prestigious state that we can feel good about.

Explaining the Pattern of Government Subsidies

The prestige argument helps explain the observed pattern of government subsidies, which the other externality arguments cannot do. Prestige aside, arguably the most serious externalities are for electronically reproducible popular culture. The ease of unauthorized copying, even prior to online Internet file trading, means that musical artists do not reap the full reward of their labors. While the Internet may open up more opportunities than it will foreclose (see chapter 4), a market imperfection is likely to remain. Musical artists create some social benefits that they do not get paid for, and too few recordings are produced, relative to an optimum. At least in principle, a subsidy to popular music recordings could remedy this inefficiency.

High culture is less likely to confront an externality of this kind. Many forms of high culture are performed live, which makes them harder to reproduce on a mass scale. While bootlegging of live classical performances is a common practice, it has little effect on the incomes of performers. If Maria Callas performances had not been systematically bootlegged by her fans, her income would not have been substantially higher and in fact might have been lower, owing to diminished popularity. The heavy metal group Metallica, however, did lose business to file sharing. Similarly, unauthorized copying of discs and videotapes appears to be a less common practice for classical than for popular music, if only because young audiences do most of the copying. Mass markets, especially for the young, present the biggest problems of copyright enforcement and thus the greatest potential for market failure.

Nonetheless we see few calls for the subsidy of heavy metal music and rap. These subsidies would yield little in the way of prestige value. Teenagers, who are usually bent on rebellion, would not feel better to know that their favored art forms received government subsidy. Older partisans of high culture would reap lower and arguably negative prestige returns from such a policy. Few people would feel better about their government. That is a central reason why most direct subsidies are for high rather than popular culture.

Subsidy advocates typically favor government support for only a very limited set of aesthetic activities. They favor subsidies to opera, art museums, and classical ballet. Some commentators have pushed the margin and called for subsidies to pop music and jazz. The Netherlands, unlike most other western European countries, has subsidized puppet theater and pantomime. The Japanese subsidize flower arranging. But few individuals favor a wide extension of subsidies beyond traditional high culture. The NEA has programs for the American folk arts, and for jazz, but these yield prestige through mechanisms of patriotism and political correctness.24

In this context we should not think of the aura of opera as somehow "causing" government subsidies to high art. Rather, common processes cause both the aura and the subsidies. People have supported the aura of high art for the same reasons that they have supported government subsidies to opera. They consider opera worthy of an aura and worthy of a subsidy.25

Why do we focus subsidies on putting art in museums? The display of paintings in the Bellagio casino, in Las Vegas, attracted more attention and visitors than most traditional art exhibits. Thomas Hart Benton remarked: "If it were left to me, I wouldn't have any museums . . . Who looks at paintings in a museum? I'd rather sell mine to saloons, bawdy houses, Kiwanis and Rotary Clubs, Chambers of Commerce--even women's clubs. People go to saloons, but never to museums." To prove the point, Benton lent his Persephone painting (now in the Kansas City Atkins Museum of Art) to nightclub owner Billy Rose, who displayed it in his club The Diamond Horseshoe, where forty-three thousand people saw it in three weeks. During a comparable time span, only two art museums at that time had higher attendance, and that was counting large numbers of student field trips. The nightclub exhibit, however, yielded little prestige.26

Subsidy advocates neglect other unwelcome implications of the externalities argument. If "good" artistic works have beneficial social consequences, most likely "bad" artistic works have negative social consequences. Not all art is elevating, ennobling, educational, or of enduring value. Many artworks are dull, stupid, socially irresponsible, or offensive. Nor would "quality" artworks escape criticism in this regard. A long tradition in philosophy, as exemplified by Plato's Socrates, argues that the most moving artworks tend to be the most dangerous and to hold the greatest appeal to the base parts of our souls. In The Republic Socrates speaks of "the war between philosophy and poetry" and takes the side of philosophy. Socrates raises the possibility of banning the poetry of Homer from the ideal state and teaching only those artworks that encourage virtue.

If some artworks make our society worse, the externalities argument suggests that we should discourage them. We could censor them or at least make them more costly through legal means. Once we accept the premise that the government should be involved in the arts directly, it is a short step from "good art, encourage," to "bad art, discourage." Many subsidy advocates will disavow this conclusion (note the contrast with extreme conservatives, who often favor censorship but not subsidy), but it is difficult to see why. The very case for direct subsidies requires that government can tell good art from bad with reasonable facility. Subsidy advocates presumably have discarded the premise of governmental neutrality across the arts. We need not throw rappers into the gulag, but we can imagine a variety of policies, from selective taxes to selective copyright enforcement, that would discourage "bad" art forms from competing for consumer attention. Censorship, however, is not generally conducive to prestige, at least not in twenty-first-century American liberal discourse. Proposals to control or limit bad art have not caught on.

To summarize a few of the conclusions, the prestige argument can clash with the decentralization argument. The decentralization argument calls for effective subsidies, whereas the prestige argument calls for visible subsidies. The economic version of the decentralization argument calls for general support for creativity, whereas the prestige argument calls for very particular forms of arts support. We cannot resolve these differences on a purely theoretical level, so now we turn to the history of American arts programs and arts funding. In chapter 5, armed with a better empirical sense of how things have worked, we will return to the question of what exactly should be done and why.

Return to Book Description

File created: 8/7/2007

Questions and comments to:
Princeton University Press

New Book E-mails
New In Print
PUP Blog
Princeton APPS
Sample Chapters
For Reviewers
Class Use
Recent Awards
Princeton Shorts
Freshman Reading
PUP Europe
About Us
Contact Us
PUP Home

Bookmark and Share