Public goods theory Anthony de Jasay

The notion of public good presupposes a relevant 'public', a set of persons similarly placed in some respect, such as location, language, legal status, interest or need, who all have free access to the good by the sole virtue of being members of the set. Publicness implies that, though some or all members of the public may contribute to the cost of the good, for example by paying tax, their access to it is not 'bought' by their contribution. (Public goods differ from 'club goods' in that access to the latter is open to contributors only, for example, to those who pay membership fees. However, the member may make greater use of the club good without paying more fees. Thus, while it is private intramarginally, the 'club good' is a public good at the margin in that consumption of an additional unit is virtually costless to the consumer, any increase in total cost being borne by the membership as a whole.)

The dissociation of contribution from benefit is the determining attribute of publicness. It is common to both the traditional theory of public goods and its emerging generalized version. In contrast to the traditional theory in which voluntary contribution to the cost of a public good by non-altruists is irrational, the contemporary approach treats the expected value of an individual's benefit from a public good as being probabilistically contingent on his own contribution and either smaller or larger than the contribution. By relaxing the traditional presupposition, the more general theory implies that, if the public good is valuable enough, it is potentially rational for non-altruists to contribute to it; that is, that private provision is consistent with free access.

The traditional theory

It may be held on some agreed ground (interpersonal welfare judgement, Pareto optimally or cost-benefit judgement) that producing a certain good is better than letting the necessary resources be used to some other purpose. If, however, the good cannot be provided for one member of the public without being provided for all others, the latter receive it as an externality. The contribution of anyone to the cost would benefit mainly the others, and the incremental benefit to the • ontributor would be small, typically imperceptible, and in any event '.mailer than the contribution. The problem is so defined as to nrakc this the cuse I lenee everyone is better off if the good is provided, f'ublh Hoiuh theory 1/7

luii lor each il is belter not to contribute; publie piovimnii iripnn , . mnpiil mu y contribution.

The goods in question have come to be delined by Iwn pmpi m< nmi uvalry' or 'jointness' (greater use of the good by one usei dm >. uni u « 11m ■ 11•• I« nelits available to the others) and non-excludabilily (a uirinbi i nl tin pulilii i «innot he denied access to the good). Goods having both piopeitli s 01 • pun public goods. Traditional theory recognized that there aie lew 01 no lui • 'I

H'uods that are 'purely' public. However, it sufficed that aci ess lo Ilu- good was lint or could not be made contingent upon a contribution io its itmi loi a problem of public finance and public welfare to arise, for, left lo the iuteiplas ol lice choices, the good would either not be provided at all, 01 only In •aiboptimal' quantities. The fulfilment of widely accepted optimum conditions would therefore require mutual coercion (Baumol, 1952). (Note that the prediction that the good will be 'underprovided' by the market, rather than not piovided at all, is open to the objection that, if it is irrational to contribute to its inst. it is no less irrational to contribute a little rather than a lot.)

Normative consequences

II a good is in fact in non-rivalrous joint supply, it is suboptimal to exclude anyone from access to it, since an additional beneficiary can he admitted to ai iess without any increase in cost. Conversely, if a good is non excludable, it must be provided in joint, non-rivalrous supply lo permit non dn.i iimina lory access to all members of the public.

Iliiw much of which public goods should be provided I mm resoun es made available under fiscal coercion? A necessary condition of optimal o .mini allocation is that, for every member of the public, the imaginai utility' la .I« lives from the public good should be the same as the imaginai utility yielded by every private good he consumes, and llint no resoun es aie leli unused. This, in turn, implies that the marginal rates ol translormation and ol Niihslitution between any two goods are equal to each olhct and lo then relative prices. However, an individual cannot normally effect substitutions between the resources he devotes to private and to public goods. His mar final rates of substitution are neither adjusted in nor revealed by market Itausactions. Nor do public goods have prices. A suggested solution to this Conundrum was to aggregate all individual marginal rates of substitution for a public good (relative to a numéraire private good) as it were, 'vertically', and lo consider the sum of all individuals' taxes devoted to the public good as lis price (Samuelson, 1954).

Tins, however, merely yields a formal statement of the optimum in terms ol unknowns which may or may not be capable of closer identification. Some may be defined by the preferences of a given individual, but the lattei will have no opportunity to act upon them, and therefore cannot reveal tin m in ihe course of his ordinary market transactions. Various non-market mechanisms have been suggested for inducing people to reveal their preferences for public goods, notably by ranking alternative tax-cum-public good proposals. The basic idea is that the optimal proposal will be the one that is unanimously or near-unanimously chosen (Wicksell, 1896). Taxation as voluntary exchange between the users and the provider of public goods is a common feature of these non-market mechanisms (Musgrave, 1939, 1959; Buchanan, 1968). They have a close affinity to contractarian political theory and to its manner of deriving the legitimacy of the state (Buchanan 1965, 1975).

However, intrinsic obstacles to such solutions subsist. If it is really the case that an individual's benefit from a public good is not contingent on his contribution to it, it is always best for him not to contribute, and not to honor any undertaking he may have given to the contrary. If revealing his preference for the good imposes on him a binding obligation to contribute to it accordingly, it is best for him to conceal (understate) his preference. In agreements requiring (quasi-) unanimity, his incentive to sell his veto right dearly would give rise to bargaining problems that may be intractable, or find resolution in redistributive bribes that cast doubt upon the putative merits ol' voluntary exchanges of the Wicksell-Lindahl type. Beyond these objections, the very meaning of Pareto comparisons between coerced and non-coerced levels of public goods provision is open to serious doubt.

Escape routes

Among suggested escape routes from the apparent dilemma, two have aroused wide interest. One is based on the game-theoretical finding that, when a prisoners' dilemma situation is indefinitely repeated, it is pay-off maximizing for each player to contribute as long as the others do (Axelrod, 1984) and this may make non-coerced provision of public goods consistent with self-interest (Taylor, 1976, 1987). This argument has considerable force in contexts where the members of a non-excluded public have good visibility of each other's conduct and must count on interacting with one another in similar situations in the future. Accordingly, close-knit communities of various kinds would have good chances of providing public goods for themselves without recourse to coercion. Cases of successful voluntary cooperation and self-restraint in the use of common pool resources (whose incentive structure resembles that of public goods) bear this nut (Ostrorn, 1990). Large, anonymous and amorphous groups with poor mutual visibility, however, would according to this argument presumably not resolve public goods problems, however recurrent.

The other suggested escape route is the coupling of the non-excludable public good with an excludable private good, with the latter serving as bait, as a "selective Incentive' (Olson, 1965). For the sake of getting the private good, a self-interested pet son may be induced to contribute both to it and to

I'liblh I:<»>,l\ thrill v ilit* public good. One contribution, that is, the priir ■ liaiyt d loi iln puvaii imod, must cover the cost of both goods for this solution to wink llmv i vn

11 ii. ii only possible if supplying the private gooil yield . .1 ».iipem |nnlii which can be devoted to subsidizing the public good t'oinpi ««»ulil then seek to supply the private good or its close subNtilutcs ai ti lowci pin < A siiong monopoly would be needed to defeat such attempts I low lit* inn nopoly permitting the supernormal profit could arise and pcisisi c, noi 1 Inn, ind the supposed role of 'selective incentives' correspondingly III delined

'Government failure'

taking it for granted that voluntary provision of public goods must be aborted bv 'market failure' due to non-excludability, traditional theory considers piiivision by the government, financed by taxes, as patently superior to the purported alternative of no public goods, lawlessness and a 'nasty, brutish •lid short life'. It seems to go without saying that, since only the government • un, it ought to supply public goods: 'the only question which arises is w lielhcr the benefits are worth the costs' (Hayek, 1960, p. 222). If this is the niily question (and in a trivial sense there is indeed no other), it is a large one I here is no agreed way of answering it. How much of which goods are to be piovided are matters decided in complex political processes, For well known tendons explored in social choice theory, it is highly problematic al to Ink 1 the outcomes of these processes as the 'true', credible, reliable expressions ol society's preferences'. Yet, if available social choice mechanism* aie not in I» trusted, what other indications should one follow? On what giotimK • an iinr affirm that the benefit of a public good is not worth Its 1 ohIV

Any political process offers opportunities for a winning loalltlon • *• ills lilhtile the benefits and costs of public goods asymmetrically, skewing iln 'product mix' to favor its own interests and tastes, and making the ■ osis lull mote heavily on the losing coalition. 'The benefits are woitli the cost' to the vs inuers, but perhaps not to the losers. Some or all of this ledistributlve eflect Is 1 oncealed from view, and some of it may be unintentional Nclthei feature •dualId commend it.

Secondly, there is a presumption that the political process is systematically biased to overprovide the good relative to some putative Pareto optimum Hie bias may arise from 'fiscal illusion' - people vole for incremental ex pendilure without perceiving a connection between it and their own incic mental taxes - from asymmetry between those who vote for public goods and Iho-.e who are made to pay for them, and from the ability of single issue 1'ioups to press for and obtain expenditure on some special public good, ol interest to the single-issue 'public', by strategic voting or 'log-rolling' A* Hayek put it, a public good must satisfy 'collective wants of the commiiuiiv .1. a whole and not merely ... of particular groups' (Hayek, 19/H, p III), but this merely begs the question of how we tell what 'the community as a whole' wants, and how we make the political process deliver it (rather than some dubious aggregation of several particular wants).

Thirdly, there is a tendency to systematic overprovision of public goods as long as their production responds to their use, consumption, or to 'need'. The marginal cost to the user is imperceptible, hence he has no inducement to use the good economically; a bias towards wasteful use is probably inevitable and is confirmed by experience, notably in publicly financed health care and education. On these and similar grounds, the provision of public goods by compulsory contributions is surrounded by 'government failure' that takes the place of 'market failure'. However, the balance of the argument might still lean in favour of compulsory provision of at least some goods if voluntary provision were, as the traditional view used to hold, clearly infeasible.

Generalizing the theory

The received view of public goods is a special theory in that it rests on three particular and demanding assumptions: (a) the goods in question are nonexcludable; (b) they are in joint supply; and (c) an individual's contribution to the cost of providing the good increases his benefit from it by less than his cost. Each assumption can be usefully relaxed, rendering the theory capable of explaining a wider range of phenomena.

Excludability is not a binary, yes-or-no relation, but a matter of degree, best seen as a continuous function of the cost of excluding access to the good. This exclusion cost is, in fact, the cost of protecting (enforcing) full property rights in the good (Demsetz, 1964). Part of the enforcement may itself be supplied as a public good (through the maintenance of law and order), while some of it is normally assumed by the holder of the right, the owner. By incurring the cost, the owner (typically the original producer or the reseller) ensures that access to the good is contingent on his consent, which he sells at a price. The exclusion cost can vary from the very low (such as measures to prevent shoplifting) to the prohibitively high. Some exclusion costs are 'objective', a matter of technology and logistics; others may be subjective (for example, (he pity, shame and social odium involved in refusing medical treatment to the indigent, or in excluding poor children from school, instead of treating such categories of people as a non-excluded public). A political decision to provide a good publicly may be due to its high 'objective' exclusion cost which would prevent profitable production for sale, or to a high imputed, 'subjective' cunt, such as social disapproval, unrest or the risk of electoral defeat.

In principle, it seems possible to provide all goods freely to a defined community, subject to a budget constraint. This is actually the case in many families where niluoi children gel 'all they want within reason', and it was

0 0

Post a comment