Public Goods as A Category Mistake:

In one of the classics of twentieth century philosophy, The Concept of Mind, Gilbert Ryle argued that many people find difficulty in understanding the differences of logical status between related concepts, and that the logical categories in terms of which analysis takes place are often wrongly selected. To say that mental processes occur does not mean the same sort of thing as saying that physical processes occur and

Ryle's book focuses on the Cartesian category mistake in which mind and brain are confused and are allocated to the same logical category. However, this is only one, startling, example of the genre. The problem is a general intellectual one, and Ryle gives many other examples.

It is my contention that the concept of Public Goods, as analysed by economists, involves such a category mistake, and that this mistake renders many of their analyses worthless. One might apply Ryle's remark, above, to the problem and say that the belief that there is a polar opposition between Public Goods and Private Goods is the belief that they are terms of the same logical type.

But they are not.

Because there is an assumption that private and public goods are terms of the same logical type, it is also assumed that the techniques for analysing the one will be appropriate for analysing the other. When Samuelson started the modern analysis of public goods rolling he did so from the perspective of an intellectual framework that had been developed for private goods.

Exactly what is a public good is a matter of some debate among economists. Several criteria have been advanced, including non-rivalness, non-excludability, and the impossibility of rejection. Different approaches put primacy on different criteria, or combinations of the criteria. This leads to disagreement over whether things like bridges, TV signals, and air, etc., count as public goods. This in turn leads some sceptics (e.g. Malkin and Wildavsky, Journal of Theoretical Politics, vol 3 no 4 pp 355-378) to argue that the distinction between public and private goods should be abandoned. It is my argument that public and private goods are concepts of two different logical types, in the same way that Ryle claims that the mind and the body are.

I do not wish to become bogged down in the arguments about whether this or that is a public good, rather I would like to focus on a couple of examples that tend to be accepted by economists, namely defence and lighthouses. Of course, public goods are only public goods for people in particular contexts - e.g. the Lizard Lighthouse in the English Channel has no relevance for mariners in the Sea of Japan, and the UK's defence has in principle little relevance (depending on the nature of the attack) for, say, Aruba. Also, public goods are only so in practice. If non-excludability is a key criterion, much depends on the state of the prevailing technology - defence could cease to be a public good in the traditional sense if a sci-fi style force shield is invented that could protect people on an individual basis (though see Olson's point, below). However, let us consider these standard examples of defence and lighthouses and take it from there.

Defence

What is Defence? Is it a gun, a tank, or a bomb? Is it a lot of guns, tanks, or bombs? A gun is a private good. It may be owned by an individual, or an organisation, and others may be excluded from its use. An individual or a state may own many such private goods. Mere ownership by the state does not transform the gun from a private good into a public good in the sense of 'public good' that is being investigated here. It is, of course possible to own these things without there being any defensive relevance if, for example, one was running a museum perhaps devoted to illustrating how the design of guns developed historically. Or maybe you just happen to have one lying around up in the loft. Thus, there is more to defence than merely possessing many private goods, whether they are guns, tanks, missiles, bombs, or whatever.

Of course, strictly speaking, these private goods may be used both to fend off another's attack and to attack another, and their effectiveness may arise both from actual use and from the threat of use. In the real world 'defence' tends to be used to cover both the fending off and the offensive activities. As I write this the Departments of Defen(ce/se) of Britain and the USA are preparing to dump shedloads of private goods on Sadam Hussein in Iraq, and it is the latter who will need to be doing the fending off. I suppose it counts as a triumph of public relations that nations invariably have Departments of Defence but not Departments of Attack. Departments of War are usually just short term organisations for efficiently prosecuting wars already in the pipeline. In the theoretical world, however, one would need to make a distinction between the two activities for the purpose of a definition.

Not that I need to provide a complete definition of any given public good for the purposes of my argument - all I need to do is say enough about its nature to make it clear that it falls into a different logical category from that of private goods.

So what is the relationship between these private goods and defence? The first thing to note is that, in principle, defence relates to the nature of the attack. One set of private goods may be effective in fending off one kind of offensive private good but not another. In other words they may provide defence or they may not. Actual defence is only provided when there is actual attack, otherwise the defence is potential. Let us focus on the actual. What it is that they provide can be put in various ways depending on the circumstances. In general, however, actual defence can be seen as successfully causing the attack to fail. This is clearly a public good according to the mainstream criteria. It is non-excludable for the people for whom the defence is relevant, no person can die from a failed attack. It is also non-rival for them - one person's absence of death does not affect another person's absence of death. And it is impossible for anyone to abstain from having an absence of death caused by the attack. What we have here is a concept that is logically quite different from a gun or a bomb, just as the mind is different from the brain.

One implication here is that defence is a binary concept. If a system of private goods is put into place with the intention of providing defence for a particular area against a particular kind of attack, then either it is successful, if everyone survives, or it isn't, if say only 75% survive. Of course, if the intention was only to protect 75% of the population, then it is successful. In practice, much depends on the objective of the defence, but once that has been formulated with sufficient clarity, it becomes obvious that the concept is binary. The private goods, while individually discrete, may be supplied in more or less quantity. This clearly has implications for anlaytical approaches.

Of course, in the real world things are often, or even usually, uncertain. Politicians often don't know what quantity of private goods to provide for a given contingency. Often they cannot, or will not, afford whatever is necessary, and take risks, sometimes pretending that what will only protect some will actually protect all. In practice, all manner of uncertainties, imprecisions, and obfuscations will exist. But this does not alter the underlying theoretical analysis that the two concepts have a logically different status, and that two quite different sorts of analyses are required - just as in chess tactical principles are one thing and the chess pieces and their rules of movement are another [ Ryle, p. 78 ].

Thus, 'defence' is the concept that represents the potentialities, capabilities, etc., of the various private goods involved. It is the way in which we look at how the particular sorts of private goods (guns, bombs, etc.) are relevant to certain situations in which there is a potentiality for getting killed or otherwise seriously inconvenienced.

Lighthouses

A lighthouse is not a complicated concept: it is usually a tall-ish structure, made of bricks, stone or metal, etc., with some sort of light on top. In itself it is a private good. There is nothing remotely "public good" about it. However, the whole point of the lighthouse is to provide a navigational facility, and it is this navigational facility that constitutes the public good. The facility takes various forms and one can describe it in various ways, but basically it is a warning to mariners of dangerous waters and/or a kind of directional indicator that is especially valuable at night.

At night, if the light is on, the navigational facility is available, if it is off it isn't. During the day the structure itself also acts as a visible warning. If the facility is available, it is available to all mariners within the lighthouse's range of visibility; one ship taking note of the existence of dangerous rocks doesn't affect another's ability to heed the warning, and a mariner receives the warning whether he likes it or not. Of course, what he does about it is another matter altogether.

It should be clear that, again, we have an abstract binary concept that is the functional manifestation of a private good. There are not two goods here: as in Ryle's mind/body example, these are not two different species of economic goods. It doesn't make much sense to say that someone provided a lighthouse and a navigational facility.

Other examples

From the argument offered above it should be clear that analysis of further examples, whether bridges, broadcasting, circuses, or whatever, the same principle will apply: there are not two distinct goods at stake in the argument. Thus there is little point in becoming enmeshed in the complexities of arguments about whether X is a private or public good. Moreover, even if it is possible to charge for, and thus exclude people from, say, a circus, the entertainment remains a public good for those who gain entrance to the tent. As Mancur Olson says:

Conclusion

Much of the interest in the public good/private good debate has been due to the claim that the market cannot, or cannot adequately, supply public goods. 'The market' here usually means private individuals or firms, in contrast to the government, whose function is seen as supplying the public goods. However, it should be clear from the above analysis that this is a false dichotomy. They are not separate kinds of goods, the one (public) is a way of looking at the function of the other (private), just as the concept of mind is a way of looking at the behaviour of the body.

It should also be clear from this that it is unnecessary to ask questions such as whether governments (being public goods) can themselves provide public goods. It's a non-problem. Perhaps one conclusion here is that public goods should be a subject for political analysis: economists should stick to private goods.

Public Goods: How Economists Misunderstand Public Goods

© Richard Kimber 2003