Crooked Timber has got a guest-poster, Kimberley Morgan, who has been writing interestingly on the welfare state, and in particular, what in the jargon are called work-life balance issues, although having said that, I'm about to criticize the way she framed one of the issues she's posted on. Writing about child benefit, Morgan argued that children are public goods, an economic term of art meaning that they provide benefits which are non-rivalrous and non-excludable in the sense that they don't get used up when one person utilizes them and we cannot(easily) exclude people from utilizing the benefits. Clean air would be an obvious case of a public good: everyone breathes it, it doesn't make any difference how many other people have breated it, and we can't, without violating various fundamental rights, easily prevent people from breathing it. Children, she argues, are public goods because they provide a future workforce which will fund our retirements, and thus, because public goods, because it is difficult to assign property rights for them and hence get trades towards an equilibrium, are not best distributed through a market mechanism, child benefit makes sense: we are subsidizing production of children to the level which is socially optimal.
This may well be true: given state-funded old age pensions, someone needs to be earning a taxable income to fund our retirements, and even in their absence, if we don't want the economy to collapse due to lack of labour when we retire, children are probably needed. However, it misses the point of child benefits completely. The question is rather whether we could justify the costs that would appear in the absence of a system of child benefits to those who would bear them, and the costs of having such a system to those who would bear them: whether parents and children would have a legitimate complaint about a distribution of economic goods which did not include state-guaranteed child benefits, and whether those who would pay for those benefits would have a legitimate complaint about the costs paying for them imposed on them. Talking about them in terms of public goods is quasi-utilitarian: rather than optimizing across society, as talk of public goods does, because of its focus on the equilibrium across the aggregate, we should be thinking in terms of whether any particular set of individuals, considered as individuals, have any good reason to reject the policy in question and the sacrifices it demands of them.
A public good, if it is a true public good, will be to the benefit of all, and state intervention in a market to correct the market's failure to allocate it efficiently - because of the problems of applying property rights to it - should theoretically force the market to allocate as it would do were it able to work efficiently: the cost of the provision of the last unit of the good should equal the benefit of the provision of the last unit of the good to whoever benefits from it. Thus, if children are being considered as public goods by virtue of their capacity to provide for our retirement, it should be the case that we should fund their production until the costs of doing so are, at the margins, equal to the benefits to us, at the margins, of having them there to fund our retirements. I have absolutely no idea at what level this might be: it would depend on the structure and efficiency of capital markets, which determine the effectiveness of saving, and on the structure and efficiency of labour markets, which would determine the level of income at which it would be efficient to tax these children when they are working. It is also incredibly vulnerable to the retort that, for the childless and wealthy, children are much less of a public good than for the poor: these people do not need the next generation to provide for their retirement, since they were able to save enough for themselves, and so the public good falls back solely on the effects of having a next generation to form a labour force at all, which requires fewer people, and so less child benefit, since fewer children need to be produced.
Instead of attempting to use a quasi-moral economic framework to justify public policy, we should be asking, is what is given up by those who fund child benefit more or less than what is given up by those who recieve it, were they not to recieve it, in general terms. It seems fairly clear to me that generally it is less: ensuring that children do not grow up in poverty by taxing income at the margin over a certain, comfortable, level, does not seem to me to grant a benefit to those children smaller than the cost imposed on those who pay for it. Equally, legislation to ensure that parents can get time off work, and return to their jobs relatively easily afterwards, when they have small children, seems to impose a cost on businesses, and through them, their shareholders and perhaps the economy at large, but, at a sensible level, this costs are reasonable to impose to avoid impose a cost of seriously damaging your career, or your relationship with your children, should you have children.
Neither is this vulnerable to the kind of retort mentioned above: it is immaterial whether or not one individual in particular bears costs that they do not - in financial terms at least - recieve back at some point, because the question is whether, given the benefits recieved by the beneficiaries, the cost imposed on them is justifiable. These costs should not be thought of in financial terms, where a loss of a couple of thousand pounds for a stock-broker is equivalent to the gain for a working parent on a much lower income. The costs should be thought of in terms of what Scanlon calls 'generic reasons', the reasons we all, generically, have: wanting to live in a pleasant environment, to have some control over our lives, to be able to have a little fun in our lives, to be able to form and sustain lasting relationships with others, and so on, many of which are significantly determined by our access to financial resources. The benefit to the poor of transfers from the wealthy to the poor in terms of these purposes is significantly greater, generically, than their costs, because money, generically, has decreasing marginal returns to these purposes beyond a certain point. Hence they are justified.
Nor is it vulnerable to the libertarian 'it's mine, and you can't take it' retort: the question of the extent of property rights is to be decided by exactly the same procedure. Property rights are a social institution like any other, and require justification. Patently, absolute property rights, to the exclusion of all taxation, are utterly unjustified: they tend to become concentrated, and those left with few or none, tend to do very badly. No millionaire can justify to the starving poor on the street the couple of hundred thousand that allows them to employ an extra maid, and add another wing to their mansion: the benefits to them of that particular bit of their income are, in moral terms, vastly outweighed by the benefits that others could reap from it. This is not to say that the efficiency of particular policies does not need to be considered: there certainly are tax rates at which, even though the beneificaries of the redistribution do gain more than those who pay for it lose, the disincentive effects to society as a whole make up for the difference. This is included in the moral reasoning though: the policy is unjustifiable because everyone suffers as a result of it.
I recently wrote a piece arguing that Stumbling and Mumbling's use of positivist economics seemed a little odd, given their use of the usually anti-market managerialist critique of post-Enlightenment moral and political philosophy. Whatever else one might think of Stumbling and Mumbling, this is a classic example of what I take it that the managerialist critique is supposed to skewer: efficiency, in the sense given above, as a given and unquestioned end of public policy, effectively morally contentless, working to justify public policy (I'm not saying that markets must be like this: Nozick's is a moral theory, even if it is wrong, and Dworkin's envy-free auction is supposed to give everyone equal entrance into a market, which then justifies itself, because the cost to each of forgoing what they do not have is matched by the benefit to each to having what they do have, which justifies itself precisely because the costs and benefits are equal, because of the initial envy-free distribution, although it's wrong as well).