Brian Barry (Reference Barry1965, pp. 112–15) once suggested that the concept of desert flourishes in a liberal society in which each person's worth can be ascertained by the market, a society like that of the late nineteenth century. He went onto to suggest that public policy in the twentieth century saw a ‘revolt against desert’, exemplified in a welfare state that catered for the underserving poor. By contrast, Hayek (Reference Hayek1976, pp. 70–3) famously saw the idea of equating merit with returns in a market economy as a mirage. He saw the revolt against desert not as an abandonment of a liberal market order, but as a necessary condition for understanding how such an order could work well. The extremes touch: either way, desert is seen as a thing of the past.

Oliver (Reference Oliver2021) suggests a counter-revolution to this revolt against desert. He is motivated by a concern that public perceptions of the legitimacy of income transfer systems depend upon their respecting considerations of desert. The paying public are entitled to demand that those at the bottom of the income distribution genuinely deserve assistance. Although Oliver does not draw this implication, a demand to incorporate considerations of desert into income transfer programmes might come as much from net beneficiaries as from net contributors, since the beneficiaries who constituted the respectable poor might regard themselves as deserving, unlike those whose poverty was not due to unavoidable misfortune but to the lack of thrift.

Oliver contrasts public perceptions with what he calls Rawls's ‘transcendental proposition’, the principle of just distribution derivable from the original position. Oliver suggests that the view from the original position will lead decision makers to focus solely on the least well-off independently of their deserts, since behind the veil of ignorance those decision makers would focus their minds on how they would wish to be treated if they were in unfortunate circumstances. However, so the argument continues, this view from the original position is unaligned with the world in which we live, and the requirements of people living together will allow for knowledge of personal circumstances and, hence, for considerations of deservingness.

Without getting side-tracked too much into questions of Rawlsian exegesis, there is a way of understanding the supposed reasoning of the parties in the original position that is closer than this interpretation suggests to the position that Oliver himself adopts in his article. Rawls came to regret his earlier presentation of the theory, which made it seems as though deriving his principles of economic justice depended upon implausibly risk-averse assumptions being ascribed to the parties behind the veil of ignorance about where they would end up in the income distribution (see, e.g., Rawls, Reference Rawls and Kelly2001, p. 97). Instead, the focus should be not solely on the least well-off, but rather than the worst that could happen to any citizen of a society, which is why Rawls came to think that the argument for equal liberty via the fundamental constitutional freedoms was more forceful than the argument for the difference principle (for an exposition, see Weale, Reference Weale2020a, pp. 194–97).

Where Rawls and Oliver do part company, however, is in relation to Rawls's more general arguments for rejecting the principle of desert, arguments that are independent of the construction of the original position. Rawls asserts that a good character depends upon the family in which someone grows up and that it is a matter of mere contingency where any one person ends up in the natural distribution of skills and talents. Even the disposition to effort is a given (Rawls, Reference Rawls1999, pp. 273–77). Some (e.g., Cohen & Nagel, Reference Cohen, Nagel and Rawls2009, pp. 18–19 and Nelson, Reference Nelson2019, but see Weale, Reference Weale2020b) regard these views as an echo of Rawls's earlier neo-orthodox Protestant theology in which merit is rejected as a ground of just treatment on the grounds that even the most upright person would stand condemned by the Word of God (Rawls, Reference Rawls and Nagel (ed.)2009, p. 240).

However, we do not need to invoke these theological premisses to be sceptical of the role a considerations of desert might play in the theory of justice. Rawls's scepticism about desert was anticipated by Sidgwick in his discussion of the ‘common sense’ idea of justice in The Method of Ethics (Sidgwick, Reference Sidgwick1907: Book III, Chapter V). The whole of Book III is taken up with trying to clarify a range of common sense or intuitive ideas of ethics, including justice, in relation to which Sidgwick considers how far desert can be a sufficient reason for assessing the justice of a distribution. He points out that we cannot easily reward someone as deserving because the utility of their service may have been developed by favourable conditions of life or by a good education. Good dispositions and habits are partly inherited and partly due to the care of parents and teachers. We cannot separate in practice that the part of a person's actions that are due to free choice from that part which is due to the original gift of nature. These criticisms from Sidgwick do not stem from a transcendent point of view, but from the difficulties associated with making determinate the idea of desert in any detailed practical implementation.

It might be argued that a philosophical inability to provide a clear conceptual account of desert, of the sort that Sidgwick explored, is not a barrier to use the principle in the administration of public policy. After all, it has proved philosophically impossible to offer an adequate account of the grounds and conditions of legitimate punishment, but the practice continues to exist, with judges passing sentence every day on convicted criminals in a widely accepted practice. If judges can sentence criminals, on the grounds that they deserve their punishment, could not public officials make determinations of an entitlement to income transfers or other forms of social assistance taking into account the extent to which beneficiaries were deserving? Does Sidgwick's conceptual indeterminacy really have any practical relevance?

This is not a new question. Indeed, the issue was discussed extensively in the debates in the first part of the twentieth century around the future of the Poor Law, which at the time provided a legislative and administrative framework for dealing with questions of poverty and destitution, a framework that originally went back to Elizabethan England and into which was built the distinction between the deserving and the undeserving poor. The turning point in that debate, though for specific historical and political reasons it took some decades to have an effect, was the 1909 Minority Report of the Poor Law Commission (Webb & Webb, Reference Webb and Webb1909a, Reference Webb and Webb1909b). The authors of the Minority Report were specifically concerned with the lack of uniformity in the policies of the Poor Law Unions, the collections of parishes responsible for the administration of the law, a problem on which a number of previous commissions had commented. In the present context, one particularly interesting chapter of Part 1, the Report is concerned with the administration of ‘outdoor relief’, social assistance that did not require entry into the workhouse on the part of the recipient, but which was given to the non-able-bodied poor in the form of payments of one kind or another.

In discussing outdoor relief, the Minority Report identified numerous inconsistencies in the way the Unions took into account the virtue of thrift (Webb & Webb, Reference Webb and Webb1909a, p. 33). On almost any account of the way in which considerations of desert ought to figure in decisions on social assistance, thrift ought to be grounds for beneficial treatment. After all, if the poor demonstrate thrift, then they are deserving, not undeserving. Yet, as the Minority Report notes, ‘the occupation of a small holding, the holding of an allotment, the keeping of a cow or a donkey, or the possession of poultry, is, in some Unions, actually a cause of disqualification’. A large part of the Report is taken up with detailing the ‘Babel of principles’ (Webb & Webb, Reference Webb and Webb1909a, p. 37) that different Unions adopted when dealing with similar types of cases of claims for assistance. Of particular note are the inconsistencies in dealing with mothers and children. In some Unions, for example, deserted wives are treated as widows and in others not. Widows with children may or may not be treated differently from widows with illegitimate children and so on. The implication here is a simple one. It is one thing to have intuitive convictions about how deserving someone might be in a particular set of circumstances; it is another to design an administrative system that can make moral deservingness central to the determination of assistance without generating inequity and so injustice. In short, the practical corollary to Sidgwick's indeterminacy about the proper grounds on which desert might be judged is an administrative system will display the lack of uniformity and wayward discretion.

If the Sidgwick–Rawls critique of desert is telling and the problems of administrative design overwhelming, what would that mean in public policy terms? In particular, would these considerations imply, for example, a policy of unconditional income transfers, in which all citizens were guaranteed a basic income? One of the concerns behind Oliver's interest in desert is a worry about the erosion of public support for transfer programmes if those programmes are associated with benefits to the work-shy. Of course, there may be nothing morally wrong with the work-shy. After all, we have it on good authority that in our lives we should consider the lilies of the field ‘who toil not neither do they spin’, but still outdo Solomon in all his glory. Yet, public sentiment may not be up to this demanding insight and consequently resist unconditional transfers. If the rejection of desert implies the unconditional transfers of a basic income scheme, how far would this be compatible with public support for any scheme of redistribution? Fortunately, we do not have to face this possibility since we do not need to invoke desert to resist the inference to unconditional transfers. Questions about the morality of diligent work habits among particular individuals can be put aside. What is at issue are in the design of income transfer programmes are not questions of virtuous character but the need for reliability in the contributions made by those benefiting from a common scheme of insurance against ill-fortune. The dispositions of reciprocity that Oliver sees as deeply rooted in human nature – to cooperate with those who cooperate with us – underlie the ability to construct institutions and practices that share the risks inherent in the life cycle. As Nicholas Barr (Reference Barr2001, p. 1) has pointed out, even if we were all middle class (and therefore by definition all equally deserving), we should still need the institutions of the welfare state. But those institutions cannot flourish if there is free-riding by those able to work and pay contributions but not willing to do so. Making benefit depend upon contribution, suitably modified to cover unpaid caring responsibilities, is a way of ensuring the integrity of the income transfer institutions on which the vast majority of people rely at some time or another in their lives. Prudent design against free-riding and not considerations of desert is all we need.