IX. Governmental Provision of Public Goods
We have concluded that a society of people with rational and moderately virtuous dispositions will provide significant amounts of public goods, but not at optimal levels. How should a society respond to this failure? We may divide possible responses into four broad categories: i) do nothing, ii) convert the good into a private commodity (e.g., parcel out land held in common) iii) regulate people's behavior (e.g., require pollution control devices on automobiles) iv) tax and provide (e.g. tax incomes to pay for national defense). Different responses are appropriate for different kinds of public goods or bads, different local histories and traditions, different technological and organizational capacities. Here I cannot hope to comprehensively discuss all four types of governmental responses to the failure of private provision. Instead, I would like to focus exclusively on the arguments for doing nothing.
Some arguments against governmental provision
Inadequate voluntary provision does not immediately imply that states should assume the responsibility of providing or protecting every public good. The state may do an even worse job than private, partially self-interested individuals acting on their own. There are several reasons to think the government will be a relatively clumsy provider of public goods. Bureaucracies do not have to compete, so they have few incentives to be efficient. Bureaucrats tend to avoid big, but potentially rewarding risks, and are reluctant innovators, since there is less latitude between the consequences of success and failure (salaries vary over a smaller range). Without risk-taking there is dampened innovation and limited experimentation, leading to smaller expected returns than is socially desired. Finally, the threat of bankruptcy does not discipline governmental administration.
By comparison, diverse and independent not-for-profit organizations avoid most of these shortcomings. While there can only be one federal government and one local government to serve a single locality, there are thousands of nonprofit organizations, some of them with overlapping missions. Hence, nonprofit organizations are able to operate more like a firm in the market place, competing with each other to satisfy donor demand. If they take chances they can grow or dwindle quickly. Because nonprofit organizations, like profit making companies, grow at different rates, attain different sizes and have a natural diversity of characteristics, they are better at responding sensitively and quickly to donor preferences and client needs in a complex society.
Of course, salary ranges are even smaller in the nonprofit sector than in government agencies. However, they still seem to lead government agencies in innovation and creativity. Compare, for example, the prevalence of these qualities within nonprofit environmental groups and environmental government agencies of equal size. Although now run by many state agencies, it took the Nature Conservancy to conceive of the Natural Heritage Program, a project to catalog and survey a region's species and store this information in a computerized database.
Finally, nonprofit organizations are often painfully aware of the threat of bankruptcy.
Another consequence of governmental intervention is the discouragement of technological innovation. State provision of a public good removes the incentive to discover new technologies that allow for the affordable exclusion of free-riders. For example, commercial free t.v. broadcasting is a public good, but with the invention of cable t.v., the transmission of the same content can take the form of a private good (though today, still not as affordable as open air broadcasting). As toll booths become more efficient, the cost of restricting access to a roadway declines.
Some economists would like to add still another reason: the cost of "rent-seeking" ("rent" here meaning "economic rent", which is different from the rent a tenant pays her landlord). They suggest that the lure of large governmental subsidies encourages potential beneficiaries to spend resources to secure advantages for themselves. Many believe that spending resources in a competitive struggle to secure governmental favors is wasteful. However, one should avoid too hastily condemning rent-seeking behavior without thinking carefully about the effects. If an industry or interest group spends money trying to persuade politicians to deliver some benefit, they will have to perform valuable work to succeed: they will have to provide politicians with information and arguments the politicians find convincing. In the competition for favors, they are forced to contribute in a useful way to public discourse.
Note that an obvious example of rent-seeking behavior--the bribery of politicians--in addition to being illegal, does not lead to a Pareto inferior outcome. Bribes which secure governmental favors simply transfer money from some individuals to others. The redistribution of wealth is not, per se, inefficient. If someone wishes to claim that rent-seeking is inefficient, he must show that the kind of redistribution involved is inefficient. He might, for example, try to argue that the redistribution results in greater inequality and this inequality (because of the declining marginal value of wealth) is inefficient. But this line of reasoning seems rather implausible.
The cost of fighting corruption, on the other hand, like the cost of locks and alarms, is a true welfare waste. However, it is unclear that corruption is more expensive to control in government than in the private sector. One would think corruption is easier to fight in government than in private enterprize, at least in the United States, where news services have greater legal access to the activities of public servants than private business managers and tend to scrutinize the former more intensely than the latter.
One might argue that what would constitute bribery in government would simply be bidding for services in a free market. If a firm wants to gain a governmental contract, it bribes an official. If the firm wants to gain a contract with a corporation, it sweetens a legitimate offer. Private markets seem to avoid the costs of enforcing and evading laws against bribery by offering legitimate and efficient alternatives.
However, it is doubtful that the costs of law enforcement and law enforcement avoidance would be any less if private firms were to provide public goods and services rather than the government. Persons with venal motivations would still be tempted to subvert the bidding or bargaining process were large private firms handing out goods or services, just as they are when the government is handing them out. Losers of (or those fearful of losing) a competitive process, whether it be public debate or a sealed bid auction for a governmental or a private contract, would be equally tempted to get what they want by bribing a private employee as by bribing a governmental official. The rent-seeker, for example, can just as easily bribe a private firm's employee to gain a share of the firm's business at an inflated price.
Accepting a bribe may have unfortunate consequences for a firm. The bribe may force the firm to pay more for a productive input or to buy an input of inferior quality. If significant inefficiency results, the firm's competitors may put it out of business by offering lower prices or better quality services or products. However, while this is unfortunate for the firm, the employee's interests are not identical with those of the firm. In a dynamic market with labor mobility, an employee could take the bribe and then switch to a more competitive firm. If the bribe is large enough or the employee is at the end of her career, she could simply retire.
Presumably, the amount of money firms spend on "courting" the employees of potential clients or business partners is quite significant, especially when critical deals are being negotiated. The costs to a firm do not seem to deter this legal alternative to bribery, especially when most firms engage in the practice. Why should actual bribery be any less of a temptation? True, in the long run, honest firms will operate more efficiently than firms with disloyal employees. If some firms have more virtuous employees, they will tend to drive out their corrupt competitors by being more efficient. However, because individual interests and corporate interests are divergent, the individual incentives to take a bribe will always remain. So there will not be the required differentiation of firms with respect to employee loyalty.
Given the much greater transparency and scrutiny of governmental operations, it is doubtful that, in the end, preventing corruption is any cheaper in the private than in the public sphere.
Special interests frequently use legal means of influence by donating to political action committees. Unlike overt payments of bribery, this money eventually ends up financing campaigns which contribute to public debate, even if most of the money goes to pay for pandering political advertisements. Since rent seekers in a democracy must ultimately "bribe" the electorate and since secret balloting prevents the direct sale of one's vote, the critics of rent-seeking behavior need to explain why such activity is inherently wasteful. It is not enough simply to point to lobbyists, as if their mere existence were proof of unproductive and senseless expenditures.
The apparent social cost of rent-seeking should be contrasted with the apparent social cost of marketing and advertising. Just as farmers will pay lobbyists to try to convince law makers to provide them with a subsidy, so too will farmers pay advertising agencies to try to convince consumers to buy their product. One may argue that in both instances people engage in a useful exchange of information and competing attempts at persuasion. In fact, rational debate and valuable education is more likely during the process of political rent-seeking than during the process of commercial advertising.
While we have found the rent-seeking argument wanting, there remain the previously listed reasons for favoring a decentralized system of private non-profit organizations over governmental provision.
A society might try to obtain the benefits of healthy competition while avoiding free rider hindrances to optimal funding by taxing the electorate and then purchasing goods and services from private companies or not-for-profit organizations. This still has some problems: a single agent, the government, makes one purchasing decision for everyone. Since the single agent is far removed from the millions of people it is trying to serve, the single agent is likely to do worse than if each person were to make her own purchasing decisions. Compare, for example, how the military feeds its service men and women and how restaurants in city centers feed equally large numbers of people. Or consider the U.S. federal government's failed experiment during the 1960's at constructing low-income urban housing. How much sooner would self-adhesive stamps have become available if the postal service did not have a monopoly on ordinary mail?
In general, the inefficiencies of governmental provision must be weighed against the inadequate funding which results from private arrangements. The more egoistic people are, the further revenue will fall short of optimal levels. For many public goods, such as our transportation infrastructure, the court system and a clean environment, the goods are too important and the shortfall too great to rely entirely on voluntary initiatives. However, for other goods, such as public radio and television, the shortfall may be an acceptable price to pay for the advantages of voluntarism.
We should not assume that all aspects of what, on the whole, are essential public goods deserve governmental funding. The issue is not that simple. A society may find it useful to have binding arbitration along side the court system, private security (e.g. doormen for apartment buildings) along side governmental security (city police), private support for museums together with governmental support for the arts, privately funded turnpike construction, in addition to federally funded highways, voluntary support for housing (such as Habitat for Humanity) as well as governmental subsidies (such as tax exemptions for home mortgages or Housing and Urban Development loans).
Oddly neglected is the opportunity for voluntary arrangements to supplement, or partially replace, our federally funded national defense system. As with existing voluntary schemes for other kinds of public goods, this would allow those who want more defense than the median voter to purchase extra military might, just as those voters who are especially concerned to cure cancer can (and do) contribute their own time and money to fight the disease. I see no more reason for making room for a degree of voluntarism in the provision of public goods related to education, medical care, or housing, than for the voluntary provision of national defense.
To some extent, the U.S. has in the past and does now provide for its defense using both methods. There have been war bonds and the knitting of gloves to send to soldiers fighting abroad. In pursuit of a widely shared view of American interests in the western hemisphere, Oliver North raised private donations to buy arms for insurgents attempting to overthrow the Sandinista government in Nicaragua. The U.S. now has a completely volunteer army, and even when there was a draft, the military recruited and trained volunteer officers.
However, the tradition of voluntarism with respect to national defense is extremely weak. In most countries during most historical periods, armies have usually been funded by the state or by individuals or classes of individuals with property interests large enough to justify and afford private armies. In the U.S. today there are no venerable organizations even partially devoted to supplementing military expenditures (with the possible exception of private military academies). Although we now have a volunteer army, the pay and benefits of recruits is intended to be sufficient to make service attractive independently of patriotic motives.
This absence of a tradition is surprising, in contrast to its presence with regard to other public goods, such as hospital construction, scientific research or human services.
No doubt, coordinating efforts with the military would present serious problems of secrecy, security and control. However, a private defense fund that supplements governmental appropriations would get around this difficulty. Such a fund could openly allocate money to various defense departments and projects, without compromising national security interests, just as Congress does.
One possible explanation would be that those people who are more concerned with military strength than the median voter, have disproportional political power so they are getting as much of the public good of defense as they want. This might be because those with the most to lose from a foreign invasion are those with the greatest wealth. Their wealth allows them to influence political outcomes much more than the median voter.
If this hypothesis were correct, however, private initiatives ought to increase during less hawkish administrations, when more people are discontent with the depth of cuts to military spending. If hawks sincerely wanted to financially support the U.S. military, abundant legal provisions would allow them to do so. They could pay for training at private academies or give the governmental military equipment. However, those with a large appetite for national security don't seem to compensate for political defeat in this way.
A second hypothetical explanation is that hawks are more self-interested than the median voter, so they have very little, if any, motivation to contribute voluntarily. However, one would expect that in a population as large as the U.S. there would remain at least some weak altruists whose desire for greater military strength is unsatisfied even during relatively favorable administrations.
A third explanation is that defense is a genuine public good, unlike certain human services or certain kinds of scientific research, over which there are public preferences, but which actually have private benefits. The next section explains this distinction between genuine public goods and these other goods which appear as public, but function more like private goods. What I have to say here may only become fully comprehensible after reading that discussion. I claim that people are inclined to voluntarily support public preference, private benefit goods at a higher level than genuine public goods. So if national defense is a genuine public good, while poverty relief, say, is not, my own theory about these goods explains why they are more likely to be supported voluntarily.
While it is true that national defense is close to being a pure public good (although one could possibly refuse protection to travelers abroad, or to residents whose prominence or geographic location make them identifiable and discrete targets), there are other equally pure examples in which voluntary support predominates over coerced support. For instance, we primarily fund public radio and TV broadcasts through voluntary contributions. National radio and TV is no less a genuine public good than national defense.
One might argue that public broadcasting fulfills less important or urgent needs than national defense. However, military operations and public broadcasting include a wide range of projects and services, which fall along a continuum in importance and may be ranked in priority. It's hard to believe that there is no overlap in importance between the two, that public broadcasting's most important service is not as valuable as the military's most trivial expense. If there is some overlap, then it is difficult to explain why projects or services of equal importance are not funded in similar ways.
So there remains an odd puzzle why voluntary funding for military purposes isn't more common. It seems it ought to be.
Public preference, private benefit goods
Not every good is a public good; some goods that seem to be public are only apparently so. When the good has a public aspect, but is private in other important respects, we must take care to sort out good and bad arguments for the governmental provision of the good. What follows is an attempt at such a sorting out. Some free-rider arguments may still favor governmental provision, but make additional assumptions (such as a heterogenous population) not required to justify governmental provision of bona fide public goods. The best arguments for governmental provision may not be free-rider arguments at all, but may instead appeal to very different considerations, such as basic principles of justice. I will argue that the state ought to provide certain public preference, private benefit goods, but for different reasons than those which justify state provision of pure public goods.
Let me first explain how we might mistake some goods with private benefits for public goods, why people might adequately, or more than adequately, provide these goods, and whether there are any realistic examples of such goods.
I define a "public preference, private benefit good" as a good that has two properties (which only apparently conflict). The good privately benefits someone or some thing. This benefit is excludable just as with any private good. However, some people also have preferences regarding the good and the satisfaction of those preferences is not excludable. If you satisfy one person's preference over producing or preserving the good, you must also satisfy the like preferences of other concerned people. This definition is not meant to supply necessary and sufficient conditions, but only to provide a rough characterization, which, together with examples, allows the reader to understand what I mean by a "public preference, private benefit good".
For the concept of a public preference, private benefit good to make sense, one must accept a distinction between welfare and preference. One must allow that the disposition of a good may satisfy a person's preferences, but not benefit her. If I care whether you have enough sorghum to eat, then your having enough sorghum satisfies my preference, but does not benefit me. Having enough sorghum benefits you and only you. Of course, your having enough sorghum might benefit me too, if, say, I were in love with you or emotionally affected by your hunger. I shall assume that it is possible to have an "external" or moral preference, the satisfaction of which is not a benefit.
Allowing for a preference the satisfaction of which is not a benefit to the person with the preference does have significant
consequences for our conception of human welfare, welfare maximization and welfare efficiency. It implies a departure from interpretations of welfare based strictly on human preferences and the Pareto test of efficiency. In place of these notions, I suggest that a more attractive view of welfare involves more than one aspect of human flourishing (such as the satisfaction of preferences), and may even involve more than just human flourishing. I will say more about the inadequacy of the Pareto test as a measure of optimality later.
We mistake public preference, private benefit goods for public goods because of the external preferences people have about them. Why not call the good "public" if the external preferences about them are non-excludable? Why focus on who benefits rather than who gets what they want? The best answer I can offer is to point out that "benefit" and "welfare" refer to the fundamental building blocks of a viable theory of value, a theory about what has value (different from a theory about what kinds of things values are). Preferences concern us only in so far as their satisfaction benefits a person, that is, makes her better off or improves her welfare. So in thinking about public goods our focus should be on what really matters: welfare, not preferences. Of course, one may define a technical term however one likes. These are my reasons for defining "public preference, private benefit good" and "public good" in the way that I do.
I will show that if rational agents provide public preference, private benefit goods voluntarily, they will either provide them in amounts that are adequate or excessive. Whether the level of provision is optimal or excessive will depend on transaction costs. In the instance of public preference, private benefit goods, transaction costs aid optimality rather than hinder it. This is just the opposite effect of transaction costs in the supply of genuine public goods. As we learned in Chapter Two, if there were no transaction costs, that is, if there were perfect information, free communication, and a structured bargaining environment, then people would voluntarily provide nonexcludable goods at optimal levels. They would make conditional agreements, each person agreeing to contribute to a public good on the condition that all other beneficiaries contribute their prearranged share. Each beneficiary would contribute, for otherwise she would not be able to enjoy the public good. No transaction costs would effectively remove the non-excludability property of every public good.
Transaction costs, however, affect public preference, private benefit goods very differently from genuine public goods. Without them, donors will actually over-provide these goods. We can see this best with the use of an example.
Consider two estranged parents with identical preferences. Each is indifferent to the other's welfare, but they care about the happiness of a child in joint custody. Each prefers that the child have more toys rather than fewer (no amount of toys will spoil the child), although at some point each considers an extra toy not worth the cost. For simplicity, suppose that there are only two goods available for consumption: toys for the child and clothes for the parents. If we assume rationality, preference for Nash solutions, and mutual knowledge of these facts, but no communication, then each will estimate the contribution of the other and arrive at a contribution level in which the marginal benefit of buying toys equals the marginal cost of not buying clothes.
Is the equilibrium which results efficient? It will be if we assume that the parents make an accurate assessment of the relative value of toys and clothes. Suppose, for example, that each is right to think that the welfare benefit of spending on clothes equals the welfare benefit of spending on toys. Then the parents will distribute their money in a way that maximizes welfare. Each will agree to give one third of her income to the child, which results in the child, the mother and the father each receiving one third of the combined income of the parents. (See the Appendix, "Efficient Provision Without Matching", for a proof.)
If the parents can communicate, then each will agree to buy a few more toys for the child, on the condition that the other does so as well. With the agreement, spending $1 extra buys $2 more in toys. Being indifferent to the cost to the other parent, each is happy to take advantage of the other's matching funds. If they are accurate in their assessment of the value of clothes and toys, yet neglectful of each other's welfare, then they will buy too many toys, more than would be optimal, taking everyone's welfare into account.
Of course, all of this assumes that the parents can reach an agreement about exactly how to match each other's gift to the child. If the father insists on a 2-to-1 match, the mother might reasonably hold out for a better offer. The consequent bargaining impasse could last indefinitely. A structured bargaining environment, such as the shrinking pie model with alternating offers, can remove the risk of a prolonged impasse. Bargaining is a kind of transaction cost. Here the argument has been that with transaction costs significant enough to prevent agreements, parents will adequately provide toys for their child. Without these impediments, they will oversupply the toys. Imagine a single child, a million parents--all devote to this one child--and such low communication costs that the parents are able to reach an agreement to match each other's contributions. The degree of excessive provision would bury the child in a mountain of toys (remember, however, our earlier stipulation that the child always does better with more toys).
So there are some hypothetical situations in which voluntary arrangements lead to sufficient, and sometimes even excessive, provision. Quite possibly, although I cannot testify to this fact, there are parents who are willing to split the expense of providing for their children, but each of whom is sufficiently indifferent to the costs borne by the other, that together they spend too much on their children. Are there any significant examples of a public preference, private benefit good which are easier to observe and confirm?
In significant respects, the relief of poverty is a public preference, private benefit good. Milton Friedman claims that the alleviation of poverty is a genuine public good which people will under fund if left on their own.
It can be argued that private charity is insufficient because the benefits from it accrue to people other than those who make the gifts--again, a neighborhood effect. I am distressed by the sight of poverty; I am benefitted by its alleviation; but I am benefitted equally whether I or someone else pays for its alleviation; the benefits of other people's charity therefore partly accrue to me.
Friedman goes on to say that he "accepts...this line of reasoning as justifying governmental action to alleviate poverty".
How important is it to Friedman's argument that the person want to relieve poverty in order not to have to see poor people or to personally benefit in other ways from the alleviation of poverty? It turns out to be quite crucial. If the donor's motivation is self-regarding in this way, then indeed, the relief of poverty will involve externalities that invite free riding. Voluntary provision will be inadequate. However, there may be a different motivation to relieve poverty. A person may think that her own welfare is unaffected by the poverty of others, but may nonetheless prefer the alleviation of others' poverty.
If the alleviation of poverty does not benefit the donors, then poverty relief is not in fact a public good, as I have defined that term. The only beneficiaries are those that receive assistance, and the good does not satisfy our non-excludability condition. If I give money to Jones, then I do not benefit anyone else other than Jones. Actually, we need not assume Jones is the only beneficiary of a gift to Jones, we need only assume that no other potential donors are also significant beneficiaries, and this seems a plausible assumption.
Poverty relief is not the only significant existing good that has public preference, private benefit properties. Consider certain non-human goods, such as the protection of natural systems or the pursuit of fascinating but practically useless branches of knowledge. Some people who wish to promote these goods do not wish to benefit themselves or even other individuals. They want to promote a good which they believe has autonomous value and is worth pursuing for its own sake. If they are right about the matter, then they are giving to a public preference, private benefit good. Were it not for the three considerations we will raise in a moment, these donors can expect to have the pursuit of these goods adequately funded, or--if transaction costs do not interfere--even over funded, as in the example of the estranged parents.
My argument about over provision rejects a standard of good that is reducible to the preferences of each person. I do not think that one maximizes good by maximizing the satisfaction of people's preferences. That is not to say that the standard of good is unrelated to a person's judgement about the comparative value of the non-human good. People's judgements of relative value are not necessarily wrong. Rather, I assume the opposite: that people are correct in their judgements of the value of their contributions. One useful way to measure the value of a non-human good, where there is widespread and informed agreement, is to observe how much of a person's own good she would be willing to sacrifice for the production of the non-human good. While we may use someone's preference to measure the value of a non-human good this does not commit us to the view that the good benefits the person with the preference.
To see the connection and distinction between preference for a good and its benefit consider the effect of adding people with preferences over a good. With non-human goods, having more people around who prefer them does not imply that the goods bestow a greater benefit.
Consider two populations that want to provide for a public preference, private benefit good. The circumstances are the same for both groups. Suppose people in both groups want to provide habitat for endangered insects. The more habitat they buy, the more insect species they will save or the more likely it will be that they save certain species. Assume further, that the purchased land is off limits to people, so that no one will ever see, hear or come into contact with the protected insects. Imagine that visiting the area would endanger what people wish to protect. No one will even have the psychological satisfaction of knowing that the species are protected. No one will have the psychological satisfaction because, to save the species, everyone will take a pill causing them to forget its existence. Forgetting about the species, we'll imagine, is the best way to remove the temptation to visit the protected area.
The two groups have the same number of people who are able to pay for habitat, but not the same number of people. The second group has twice the number of people as the first group, but half of them are unable to pay for habitat. Perhaps half the population in the second group has no disposable money, or live in a different country from the other half, a country which forbids them by law from contributing, or they are future unborn generations whose circumstances will prevent them from paying for the conservation measures undertaken by their forbearers. Otherwise, all the donors have the same preferences.
The donors in each group ought to contribute the same total amount to pay for habitat. The total human cost of contributing will be the same. But if the people in these two groups were to treat preserving the species as a genuine public good, then the second group would and should contribute more than the first group. There are more people in this second group with preferences about protecting the endangered species. This example demonstrates the error of regarding every good over which people share the same preferences as a public good.
If public preference, private benefit goods are private goods, then why are they not efficiently provided by perfect markets? The first fundamental theorem of welfare economics tells us that in perfect markets (no fraud, perfect information, no natural monopolies, honoring of contracts, etc.) the distribution of goods will be pareto optimal. So why isn't the distribution of public preference, private benefit goods pareto optimal?
The answer is that public preference, private benefit goods are distributed pareto optimally, but in this instance pareto optimality is a poor measure of efficiency. We need a different conception of good, such as a (possibly expanded) welfare conception (expanded to include the good of non-human objects, were we to think of non-human goods as having independent value). If we are willing to grant the distinction between preference satisfaction and welfare and if we regard the latter as more fundamental, then we will put greater faith in the notion of welfare maximization than pareto optimality. We will judge outcomes involving people according to each person's welfare rather than their preferences (although, the parents' preferences, for example, are a measure of the proper weight one should give to the child's desire for more toys).
I conclude that under certain conditions people left on their own will voluntarily supply an adequate amount of public preference, private benefit goods and may, in those circumstances where the transaction costs are sufficiently low, voluntarily supply too much of the good.
From the above theoretical discussion we should not conclude that poverty relief and other such goods are over funded. At least three factors will more than compensate for the tendency towards over funding.
First, many of these goods are partially public goods and not purely public preference, private benefit goods. For example, the external effects of poverty do affect most donors. These effects include the benefits of a more educated labor force, reduced crime, less social unrest, etc.
Second, people may simply be too selfish. While matching arrangements may provide more poverty relief, say, than what is optimal by the donor's lights, total contributions may still be less than what is appropriate. We may be glad that donors over provide because we disagree with their estimate of the value of the public preference, private benefit good. We think their evaluations of the comparative worth of a public preference, private benefit good is inadequate.
Third, and most importantly, donors are likely to have preferences of varying strengths. If the donor population is heterogeneous and donors are rational ("goods") altruists, the donors that care the most for a certain good will pay a disproportionate share of the cost. The least concerned patrons will pay nothing or carry a reduced share of the burden. To see this, imagine a game with repeated rounds of contributing. During each round, each contributor gives independently of the others, and each conjectures that the others will give what they gave during the previous round. Suppose the good is initially funded at the level that the least concerned patrons think appropriate. The most concerned patrons want funding to be higher. So they will contribute more. On the next round, the least concerned patrons learn that funding for the good is now higher than they think warranted, so they give less. To return funding to the higher lever, those that care the most must now give more. If the more passionate donors care enough, the cycle eventually reaches the point at which those who care the least give nothing and those that care the most are paying for all of the costs. The most concerned patrons have essentially bought out the least concerned patrons in order to raise the total funding level. If the more ardent contributors are not willing to do this, they must accept the level of funding dictated by the less passionate supporters. In fact, the level of funding will be dictated by the preferences of the least concerned contributors that the most concerned contributors are not willing to "buy out". Funding will be less than the amount that would be given under a scheme that required each donor to pay according to her sincere estimate of the value of a good.
Prior to this chapter, we have explored various arguments which purport to justify the voluntary provision of pure public goods. We suggested that partial altruism provides individuals with the best reason to contribute. However, because most people's altruism is partial (unlike the utilitarian's) voluntary provision will be inadequate. So we are left to conclude that even if people act on their moral beliefs, they will not adequately fund public goods. In this chapter we have considered arguments pointing in the opposite direction, which favor relieving the government entirely of its traditional responsibility in compensating for free-rider shortfall. None of these considerations demonstrate the need for a system of exclusively private provision. In the course of the chapter, we have introduced the concept of a private benefit, public preference good, which is easily mistaken for a public good, but which does not provoke underprovision for the same reasons as genuine public goods. We have suggested that it would be wrong to justify governmental provision of private benefit, public preference goods by pointing to the inadequacy of voluntary provision due to free-riding. We still defended some governmental provision, but on different grounds for private benefit, public preference goods. So the case for relying entirely on voluntary provision is weak, but not just because of free-rider liabilities. In general, wise countries will strike a balance between private and public funding and not rely too heavily on either institutional arrangement. This concludes my summary of this chapter. Now for a word about the whole project.
In the Introduction I expressed the hope that the reader would be able to learn from this dissertation about five ways to defend free-rider avoidance, which justifications succeed and which do not. I have tried to convince the reader that three of the five (egoism, social norms and group agency) are hopeless and one (fairness) is successful only under certain special, fairly restrictive and rare conditions and given certain assumptions about what counts as fair. The view I have defended most enthusiastically is that of partial altruism, which allows us to justify the levels of voluntary support for public good we actually observe around us.
In all honesty, I have not been able to iron out every crease clearly visible in a normative theory of partial altruism as regards the free-rider problem. In particular, it remains for someone else, or me on another occasion, to discover a mathematical function representing the altruist's partial concern for others which also solves the revenue ceiling problem (mentioned in Chapter VII, "Objections to Partial Altruism"). As noted earlier, one might not find a revenue ceiling sufficiently embarrassing to abandon an otherwise useful theory. I have just scratched the surface of a normative theory of partial altruism and no doubt there are many other objections I have failed to anticipate, both as the theory concerns the free-rider problem and as a normative theory more generally. Were someone to convince me to abandon the theory of partial altruism, I would look more favorably on that other consequentialist theory, utilitarianism. I might then be tempted to defend the theory against my own objections, such as the charitable trust dilemma. However, as it stands, this work, having reviewed a variety of possibilities, argues for partial altruism as the single best normative theory which can justify voluntary provision.
Back to Table of Contents page