Reconciling the liberal tradition in normative economics with the findings of behavioural economics: on J.S. Mill, libertarian paternalism and Robert Sugden’s The Community of Advantage

ABSTRACT

In The Community of Advantage, Robert Sugden reconstructs and defends an account of the liberal tradition (LT) in normative economics in the light of the findings of Behavioural Economics (BE). In this paper, the LT in neo-classical welfare economics which has roots in J.S. Mill’s thought is contrasted with Sugden’s account. Cass Sunstein and Richard Thaler’s Libertarian Paternalism (LP) arguably attempts to reconcile the LT in neo-classical welfare economics with the findings of BE and to provide it with a Millian pedigree. To the extent that it attempts this, LP is unsuccessful. While Sugden abandons core elements of traditional normative economics – notably the ‘view from nowhere’, welfarism and the preference satisfaction view of welfare – which have a Millian pedigree and advances a contractarian alternative, he successfully reconciles his reconstruction of the LT with the findings of BE and also provides it with a sound Millian pedigree.

KEYWORDS:

• Welfare economics
• liberalism
• behavioural economics
• libertarian paternalism
• contractarianism

1. Introduction

Robert Sugden is rare amongst the leading economists of his generation in that, throughout his career, he has engaged deeply with philosophy and the history of ideas. In The Community of Advantage he defends the liberal tradition in normative economics in the face of a variety of distinct forms of ‘paternalism’ motivated by the results of behavioural economics (henceforth, BE). In doing so, Sugden reconstructs the liberal tradition on the basis of a reading of John Stuart Mill which motivates the title of his book. In this paper, I contrast Sugden’s reconstruction with a characterisation of the liberal tradition in neo-classical welfare economics which is based on the writings of Kenneth Arrow. Those writings arguably also have a pedigree in J.S. Mill’s texts. This contrast illuminates some ways in which Sugden’s reconstruction differs from the neo-classical tradition. I also locate Cass Sunstein and Richard Thaler’s ‘Libertarian Paternalism’ (henceforth, LP) in relation to the relevant account of neo-classical welfare economics. This form of ‘paternalism’ is of particular interest, since it is perhaps the best-known form of what Sugden (2018, pp. 20–23) calls ‘behavioural welfare economics’ – which extends the traditional framework of welfare economics and adopts ‘fully rational’ and informed preferences which are consistent with the axioms of rational choice theory – as the basis of welfare judgements. LP justifies interference in people’s choices when they fall short of ‘full rationality’ but only when such interference does not block free choice. In The Community of Advantage Sugden (2018, chapter 4) treats LP as one of a range of forms of ‘behavioural welfare economics’. Yet, to the degree that it does not block free choice, LP is arguably consistent with elements of the liberal tradition. Furthermore, Sunstein’s defence of LP also arguably attempts to give it a pedigree in Mill’s works. LP thus arguably attempts to reconcile the liberal tradition in neo-classical normative economics with the results of BE. In as much as it does so, I evaluate the attempt. I then explore how successful Sugden is in reconciling the liberal tradition with the findings of BE.

The paper is structured as follows: in Section 1, I sketch a version of the liberal tradition in neo-classical welfare economics based on a reading of J.S. Mill and Kenneth Arrow; in Section 2, I discuss LP; in Section 3, I discuss Sugden’s work; and Section 4 concludes.

2. J.S. Mill and the liberal tradition in neo-classical welfare economics

In his statement of utilitarianism, J.S. Mill argues that: ‘[o]f two pleasures, if there is but one to which all who have experience of both give a decided preference, irrespective of any feeling of moral obligation to prefer it, that is the more desirable pleasure (Mill, 1962, p. 257). Those who have ‘experience of both’ are ‘competent judges’. Mill claims that ‘it is an unquestionable fact that those who are equally acquainted with, … both do give a marked preference to the manner of existence which employs their higher faculties’ (Mill, 1962, p. 259). The test of the judgement of the ‘competent judges’ is a version of the ‘informed preference test’ (see Crisp, 1997, p. 29) or the informed desire view of welfare which was more explicitly and fully articulated by Henry Sidgwick (1981) amongst others (see Griffin, 1986; and Railton, 1986 inter alia). Mill’s utilitarianism is consequentialist in as much as it judges the extent to which an action is right in terms of the goodness of the resulting outcome, and welfarist in as much as it restricts the relevant consequences to welfare (see Sen, 1979). Standard definitions of ‘consequentialism’ include a requirement of impartiality, so that outcomes are evaluated from ‘an impersonal standpoint which gives equal weight to the interests of everyone’ (Scheffler, 1988, p. 1).

In On Liberty, by contrast, Mill writes:

But neither one person, nor any number of persons, is warranted in saying to another human creature of ripe years, that he shall not do with his life for his own benefit what he chooses to do with it. He is the person most interested in his own well-being: the interest which any other person, except in cases of strong personal attachment, can have in it, is trifling, compared with that which he himself has; the interest which society has in him individually … is fractional, and altogether indirect; while with respect to his own feelings and circumstances, the most ordinary man has means of knowledge immeasurably surpassing those that can be possessed by anyone else. The interference of society to overrule his judgement and purposes in what only regards himself must be grounded on general presumptions which might be altogether wrong, and even if right, are as likely as not to be misapplied to individual cases. … In this department, therefore, of human affairs, Individuality has its proper field of action … in each person’s own concerns his individual spontaneity is entitled to free exercise. Considerations to aid his judgement, exhortations to strengthen his will, may well be offered to him, even obtruded on him, by others: but he himself is the final judge. (Mill, 1962, pp. 206–207)

Central to Mill’s argument is the claim that – excepting rare cases – the ‘means of knowledge’ that ‘the most ordinary man’ has of ‘his own feelings and circumstances … immeasurably surpassing’ what can be ‘possessed by anyone else’. In a discussion of this passage, Sunstein (2014, p. 7) formulates what he calls Mill’s ‘Epistemic Argument’ as follows: ‘[b]ecause individuals know their tastes and situations better than officials do, they are in the best position to identify their own ends and the best means of obtaining them’. This argument supports Mill’s Harm Principle which suggests that:

[T]he only purpose for which power can rightfully be exerted over any member of a civilised community, against his will, is to prevent harm to others. His own good, either physical or moral, is not sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinion of others it would be wise, or even right. (Mill, 1962, p. 135)

The Harm Principle rejects ‘hard paternalism’ which Gerald Dworkin defines as ‘interference with a person’s liberty of action justified by reasons referring exclusively to the welfare, good, happiness, needs, interests or values of the person being coerced’ (Dworkin, 1971, p. 108).

Are the passages in Utilitarianism on which the Millian pedigree of the informed preference test rests and the anti-paternalistic views expressed in On Liberty fatally at odds with each other? Some (like Griffin, 1986, pp. 38–39) suggest that they are not. In On Liberty Mill argues that individuals have the best information about their own situations, and are thus usually best positioned – or better positioned than others – to judge their interests. In particular, excepting special cases (such as some close family member) they are, in general, better judges than any third party. The implication is that, as a general rule, individuals are the closest approximation we have to ‘competent judges’ about their lives.

The liberal tradition in neo-classical welfare economics follows Mill in so far as it retains welfarism and consequentialism while rejecting hard paternalism. For the purposes of this paper, I take Kenneth Arrow to be a representative of this tradition. In discussing the problem of co-ordinating individual decisions in a free enterprise economy, Arrow echoes Mill in setting out his basic premises. He writes that:

In a pure model of a free enterprise world, an individual, whether consumer or producer, is the locus both of interests or tastes and of information. Each individual has his own desires, which he is expected to pursue within the constraints imposed by the economic mechanism; but in addition he is supposed to have more information about himself or at least about a particular sphere of productive and consumptive behaviour than other individuals. (Arrow, 1983, pp. 200–201)

It follows from this supposition that, consistent with an implicit version of the informed preference test, the individual is assumed to be the best judge of her own interests in the relevant domain(s) of production and consumption.

Arrow (1983) formalised some of the central results of welfare economics. These include what has become known as the ‘first fundamental welfare theorem’ which states that given standard assumptions, the outcome of maximising decisions in a competitive equilibrium is ‘Pareto efficient’ – in the sense that no reallocation of commodities can make one party better off without rendering another worse off. This theorem is central to the defence of the market in economics, and is regularly interpreted in terms of Adam Smith’s notion of the ‘invisible hand’ (e.g. Mas-Colell et al., 1995, p. 327) because the maximising choices of individuals have, as an unintended consequence, an outcome which is ‘good’ at the social level, at least in as much as it is Pareto efficient.

On this reading of Arrow and Mill, three central elements of the liberal tradition in neo-classical welfare economics, which are not always explicitly stated are:

1. Individuals are better informed than anyone else about their circumstances, and hence their preferences or desires are the most reliable basis for judging their interests or welfare (‘Epistemic Advantage’ or EA for short);

2. Under certain conditions, the (unintended) consequence of maximising choices by consumers and producers, given their constraints, is a (competitive or) market equilibrium which is Pareto-efficient in terms of welfare (‘Qualified Market Efficiency’ or QME for short).

   As a corollary of (a), I add:

3. A strong presumption against coercive intervention by third parties with a view to improving people’s well-being. (‘Anti-Hard Paternalism’ or AHP for short).¹

As a corollary of (b), the standard view is that if any of the relevant conditions (for the Pareto efficiency of the market) do not hold, there is ‘market failure’ and a prima facie case for government intervention, since the market does not produce an efficient outcome. The influence of the utilitarian tradition on neo-classical welfare economics is notable through its focus on welfare; and the evaluation of the market in terms of the outcomes of individual choices within a market. EA and AHP are, on this reading, the key elements of the Millian pedigree of the liberal tradition in neo-classical welfare economics.

What assumptions does Arrow make about rationality and motivation in making the case for the market? Arrow writes that the ‘preference pattern is assumed to possess the usual properties of a (weak) complete ordering and also suitable continuity properties’ (Arrow, 1983, p. 19). He also assumes that: ‘the desirability of a distribution X to individual j is solely dictated by the desirability to him of the commodity bundle X_j. This is the assumption that individuals act selfishly’ (Arrow, 1983, p. 20).² In another context, Arrow writes:

I do not want to rely too heavily on substituting ethics for self-interest. I think it best that the requirement of ethical behavior be confined to those circumstances where the price system breaks down … We do not wish to use up recklessly the scarce resources of ethical motivation and in any case ethically motivated behaviour may have negative value to others if the agent acts without sufficient knowledge of the situation. (Arrow, 1972, pp. 354–355)

Arrow is aware that well-meaning but inadequately informed behaviour may make matters worse. To the degree that EA is a component of the liberal tradition, the extent of government interference justified by QME must be constrained (on consequentialist grounds) in the light of the epistemic limitations of well-meaning government officials.

3. Libertarian paternalism

The findings of BE and psychology have inspired various forms of interventionist public policy. Whether the relevant forms of intervention qualify as ‘paternalistic’ depends on how one defines ‘paternalism’. The form of ‘paternalism’ to which Mill objected, on Dworkin’s reading, is ‘hard’. Some of the proposals which have emerged suggest that coercive measures are justified to further people’s welfare (see Camerer et al., 2003; and Conly, 2013). These proposals are certainly at odds with the liberal tradition in neo-classical economics, since they violate AHP. By contrast, LP classes interference in people’s choices as ‘paternalistic’ as long as it promotes their welfare; but the relevant interventions are not coercive in the sense that they do not block choice. Furthermore, the claim is that these interventions improve people’s decision making and welfare as judged by themselves. To the degree that, as a consequence, it is not a form of ‘hard paternalism’ but is nonetheless ‘paternalistic’, it is a form of ‘soft paternalism’ and it does not violate AHP. Advocates of LP might thus argue that it is consistent with one of the central tenets of the liberal tradition in neo-classical normative economics as it has been sketched above.

Sunstein and Thaler implicitly adopt a version of the informed or rational preference view of welfare (see Qizilbash, 2012, pp. 651–652). They tell us that ‘in some cases people make inferior decisions in terms of their own welfare – decisions which they would change if they had complete information, unlimited cognitive abilities, and no lack of self-control’ (Sunstein & Thaler, 2005, p. 176). They argue that in some such cases – involving status quo bias, framing, self-control, myopia and so on – BE and psychology have taught us that people systematically make choices which are either not in their best interests or irrational. To the degree that they endorse a form of informed preference view, advocates of LP might claim that it has a Millian pedigree. Furthermore, Mill himself would not have objected to some forms of ‘paternalism’. If we return to the passage from On Liberty which was cited earlier, Mill concedes that while society can intervene to aid choice, the final judgement must be made by the individual herself. He writes that: ‘considerations to aid his judgement, exhortations to strengthen his will, may well be offered to him, even obtruded on him, by others: but he himself is the final judge’. Furthermore, Mill also allows exceptions to the Harm Principle. And one of these exceptions – which Sunstein (2014, pp. 106–107) terms the ‘Bridge Exception’ – may also provide a Millian pedigree for LP. Mill tells us that:

If either a public officer or any one else saw a person attempting to cross a bridge which had been ascertained to be unsafe, and there were not time to warn him of his danger, they might seize him and turn him back, without any real infringement of his liberty; for liberty consists in doing what one desires and he does not desire to fall into the river. Nevertheless, when there is not a certainty, but only a danger of mischief, no one but the person himself can judge of the sufficiency of the motive which may prompt him to incur the risk: in this case therefore (unless he is a child, or delirious, or in some state of excitement or absorption incompatible with the full use of the reflecting faculty), he ought, I conceive, to be only warned of the danger; not forcibly prevented from exposing himself to it. (Mill, 1962, pp. 228–229)

This text arguably provides a Millian pedigree for LP because, while Mill believes that liberty consists in ‘doing what one desires’, there is a qualification that we are concerned only with desires compatible with ‘the full use of the reflecting faculty’. If we have reason to think that the desire or preference on which a person is acting is, in some sense, irrational or ill-informed there are grounds for intervention. Nonetheless, Mill generally recommends providing people with relevant information rather than coercion and so would only endorse soft paternalism. By contrast, Sunstein suggests that the argument underlying the Bridge Exception could be ‘broadened’ so that intervention might be justified in a much wider range of cases where choices do not track welfare (see Sunstein, 2014, pp. 105–110). There are at least two reasons – if one follows Sunstein’s logic – why this might be so: first, because the strong informational or rationality requirement in its implicit definition of welfare is considerably stronger than that which underlies Mill’s text, LP suggests that the findings of BE imply that there is a larger range of cases than Mill would have recognised where choice does not track welfare; and second, ‘hard paternalism’ (e.g. coercive interference or mandates) might sometimes have better consequences than the ‘soft’ variety (e.g. providing information or warnings). LP would not endorse the second of these reasons since it is an argument for going beyond soft paternalism. Furthermore, since he endorses a relatively modest rationality requirement, Mill would not accept the range of ‘soft’ paternalistic interventions advanced by LP in the light of the findings of BE.³

It is worth adding that on the strong information requirement Thaler and Sunstein implicitly endorse, given human limitations, it is not obvious that any human being could have the preferences which would, on their account, accurately track welfare (see Qizilbash, 2012, p. 652). In his version of the informed desire view, Peter Railton (1986, p. 176) also advocates a strong requirement and responds to this objection by conceding that even if nobody could actually have the relevant preferences, it provides a basis for ‘approximations’ of what is better or worse for someone. On this response, the objection is not fatal and only raises a ‘measurement problem’. In the context of LP, if one accepts this variety of response, the question is how serious the ‘measurement problem’ is and how error-prone LP is. There are, nonetheless, passages in which Thaler and Sunstein (2006, p. 259) suggest that people make better choices in some contexts (e.g. when they are choosing between ice cream flavours) than in others (e.g. when choosing between financial plans). In these passages, they are implicitly invoking some weaker requirement, which implies that choosers are more ‘competent’ in some contexts than others. The notion of ‘competence’ here might be filled out in terms of ‘having experience of, or expertise in making, certain sorts of decision and having relevant information about the objects of choice’ (Qizilbash, 2012, p. 653). In this way, LP may more plausibly claim a Millian pedigree. Yet, Sunstein and Thaler endorse no such notion. If anything, in his subsequent writings Sunstein (2014, p. 74) distances himself from Mill’s view of welfare, treating it as one amongst a number of controversial views none of which his view would endorse. Rather he suggests that paternalists should be guided by what they think would make the chooser’s life go well, on the basis of the chooser’s own view of the good life (Sunstein, 2014, p. 75). In this way, he appears to make the chooser the ultimate judge of her welfare. Yet, if the relevant judgements are supposed to be informed in the strong sense Thaler and Sunstein endorse, the relevant informed preferences would not be the judgements of any human chooser. They would be akin to the sorts of preferences Railton’s ‘ideal adviser’ – an idealised version of the agent who has ‘complete and vivid knowledge of his environment, and whose instrumental rationality is in no way defective’ (Railton, 1986, p. 184) – might have. Since nobody has such preferences, this point raises potentially serious implementation problems for the programme of LP (see Qizilbash, 2012, pp. 653–657).

How might one locate LP, and Sunstein’s defence of ‘nudging’ in relation to the account of the liberal tradition I outlined above? Clearly, LP appears to respect AHP, even if in some of his subsequent writings Sunstein (2014, pp. 4–5) challenges the Harm Principle which provides the Millian pedigree of AHP. It can also be argued that LP is not consistent with AHP to the degree that, contrary to appearances, the forms of intervention it advocates are not genuinely ‘liberty-preserving’ and can only be justified using ‘traditional paternalistic principles’ (see Grüne-Yanoff, 2012).⁴ Nonetheless, to the degree that it overtly challenges the liberal tradition in neo-classical economics, its targets must be either EA or QME. Consider EA. Sunstein (2014, p. 7) appears to take issue with Mill’s ‘Epistemic Argument’ that people are, in general, better judges of their welfare than officials to make the case for ‘nudging’ people in the direction of better choices. He claims that whether Mill’s Epistemic Argument is right ‘is largely an empirical question, and it cannot be adequately answered by introspection, rhetoric, anecdotes, or intuitions’ (Sunstein, 2014, p. 8). He also acknowledges that ‘[s]ome behavioural findings remain highly preliminary and need further testing. There is much we do not know.’ To this degree, given the current state of knowledge, he cannot himself give any definite answer to the ‘empirical question’. His chief goal seems to be to ask anti-paternalists to reconsider the plausibility of EA. Nonetheless, as I argued earlier, given that no human being can have the sorts of informed preferences LP invokes for its normative standard, it is not clear that the welfare criterion proposed by LP can be plausibly or accurately implemented. At best there is a ‘measurement problem’ which renders LP ‘error prone’. As a consequence, it is not obvious that ‘empirical work’ which attempts to establish whether interference inspired by LP leads to choices which better track welfare would be meaningful or reliable.

What about QME? Sunstein (2014, p. 16) argues – like others in this literature (e.g. Camerer et al., 2003, p. 1218) – that ‘[t]he various empirical findings allow us to identify a set of behavioural market failures, understood as market failures that complement the standard economic account and that stem from the human propensity to err’. LP here extends the scope of QME by suggesting that the conditions for Pareto efficiency of outcomes in a competitive equilibrium require ‘full rationality’. That extension would be grounded on a view of ‘full rationality’ which (I have argued) would have been unfamiliar to Mill. By contrast, the Millian texts which provide a foundation for the liberal tradition only endorse intervention in a restricted range of cases and on the basis of a much less demanding notion of rationality. For this reason, the central thrust of Mill’s arguments would not support considerable extension of interference in people’s lives and choices on the basis of QME, if its scope is extended in the light of the findings of BE. Indeed, LP is, as a consequence, not consistent with the basic thrust of Mill’s views, and to the extent that it makes any claim to a Millian pedigree, it is unsuccessful. Liberals in the Millian tradition who are dissatisfied with LP must find an alternative way of reconciling the liberal tradition with the findings of BE.

4. Sugden’s reconstruction and defence of the liberal tradition

In The Community of Advantage, Sugden has advanced an alternative. In motivating the title of his book Sugden makes J.S. Mill the spokesperson for the liberal tradition and cites selected passages from Mill’s diverse works. Yet in reconciling the liberal tradition with the findings of BE, Sugden departs in a radical way from mainstream normative economics and from some of Mill’s views which provide its pedigree. Crucial to Sugden’s account is an acceptance – based on the findings of BE – that people do not have consistent, stable or context-independent preferences. That acceptance leads Sugden (2018, chapters 1, 2 and 5) to abandon one of the central tenets of the neo-classical tradition, the (informed or actual) preference satisfaction view of welfare as a basis for evaluating advantage. Instead, Sugden offers the opportunity a person has ‘to satisfy not just those preferences that she currently has, but any preferences she might come to have’ (Sugden, 2018, p. 46) as the currency of advantage.

Sugden (2018, p. 17) also suggests that economists might consider alternatives to the consequentialist view of morality which typically focusses on the impersonal viewpoint of an ‘impartial spectator’ which he characterises as a ‘view from nowhere’. In its place, Sugden offers a ‘contractarian’ alternative which sees morals in terms of principles which potential parties to a contract might agree. Indeed, Sugden (2018, chapter 3.3) favours a specific strand in the contractarian tradition which is associated with James Buchanan. As a consequence, in characterising the liberal tradition Sugden advances arguments which are neither consequentialist nor welfarist. Mill’s Utilitarianism is, of course, one of the canonical texts in the utilitarian tradition, and on Mill’s view ‘[a]s between his own happiness and that of others, utilitarianism requires him to be as strictly impartial as a disinterested and benevolent spectator’ (Mill, 1962, p. 268). Furthermore, Mill rejected contractarianism. In On Liberty he wrote that: ‘society is not founded on a contract’ and ‘no good purpose is served by inventing a contract in order to deduce social obligations from it’ (Mill, 1962, p. 205). Indeed, to the degree that Sugden rejects the ‘view from nowhere’ and endorses contractarianism, Sugden distances himself from core elements of Mill’s thought which provide the pedigree of elements of the neo-classical tradition in normative economics. As a consequence, Sugden must provide a novel and original characterisation of the liberal tradition which is consistent with his contractarian view of normative economics. And sceptical readers may wonder whether Sugden’s claims about the Millian pedigree of his view are based on selective quotations which do not reflect the core of Mill’s thought. Is such scepticism justified?

Sugden’s characterisation of the liberal tradition endorses the following three elements: (1) ‘in a well-ordered society, co-operation for mutual benefit is a governing principle of social life’; (2) ‘a competitive market is a network of mutually beneficial transactions’; and (3) ‘a principle of non-paternalism with respect to such institutions [in which individuals co-operate for mutual benefit]: in co-operative relationships, it is for each individual to judge what counts as his or her own benefit’ (Sugden, 2018, pp. 3–4). Sugden attempts to provide a Millian pedigree for each of these elements.

As regards (2), he cites a passage where Mill contrasts mercantilism with a system of more free commerce between tradesmen across various nations and suggests that:

The feelings of rival tradesmen, prevailing among nations, overruled for centuries the general community of advantage which all commercial countries derive from the prosperity of one another: and that commercial spirit which is now one of the strongest obstacles to wars, was during a certain period of European history their principal cause. (Mill, 1852, p. 230)

Sugden’s reading of this text is that ‘Mill is telling us that if we look beneath the surface features of competition between rival businesses, we can understand a competitive market as a network of mutually advantageous relationships’ (Sugden, 2018, p. 1). This element of his characterisation of the liberal tradition requires a version of Smith’s ‘invisible hand’ argument which is supplied by QME in the neo-classical account. Sugden’s opportunity-based defence of the market (based on the ‘opportunity criterion’), originally advanced in a paper in the American Economic Review is at the core of The Community of Advantage (see Sugden, 2004; and 2018, chapters 5 and 6) and does this job. It does not require consistent, stable or context-independent preferences or even rely in any significant way on the assumption that agents have preferences.⁵ As a result, Sugden’s defence of the market supports the view ‘that a competitive market is a network of mutually beneficial transactions’ without relying on standard assumptions adopted in neo-classical welfare economics which he believes have been discredited by the findings of BE.

Next consider (3), the ‘anti-paternalist’ element of Sugden’s characterisation of the liberal tradition. Sugden (2018, p. 2) distances himself from Mill’s ‘own ideas about human well-being’ which he thinks are ‘those of a high-minded and intellectual humanist’ and which, he thinks, ‘rest on shaky foundations’. Instead, he highlights Mill’s defence of liberty and the Harm Principle. He observes, in Mill’s defence, that he ‘is scrupulous in not using his own judgements to justify paternalism’ (Sugden, 2018, p. 3). Sugden (2018, p. 2) writes that ‘[i]t is integral to Mill’s concept of ‘advantage’ that each person is the ultimate judge of what counts as advantage to him or her’. Here Sugden is relying on passages of the sort which are cited in defence of the Epistemic Argument, where the individual is taken to be the ‘final judge’ of her own interests and this element of the liberal tradition plays a similar role in his reconstruction to the roles of EA and AHP in the neo-classical account I sketched above. Nonetheless, Sugden drops Mill’s commitment to welfarism and, in his account, advantage is understood in terms of opportunity rather than well-being. As a contractarian he proposes to potential parties to the contract that it is in their interests to have more rather than less opportunity (Sugden, 2018, pp. 83–84). They must judge whether to accept that proposal.

In developing his own ‘anti-paternalist’ view, Sugden (2018, p. 46) favours a distinct definition of ‘paternalism’ due to Julian Le Grand and Bill New (2015, p. 23) according to which ‘paternalism’ is ‘government intervention’ which ‘is intended to address a failure of judgement by the individual’ and ‘to further the good of the individual’. LP and related ‘paternalistic’ proposals are out of bounds on Sugden’s account because their addressee is a ‘choice architect’ or ‘planner’ who is conceived of in standard consequentialist terms as an ‘impartial spectator’ (Sugden, 2018, pp. 17–24), rather than to ‘individuals as directors of their own lives’ (Sugden, 2018, p. 49) who are the potential contracting parties to whom proposals should be addressed. On his account, LP and related proposals thus lack a valid addressee. This argument supplies what Sugden requires: ‘a principle of non-paternalism with respect to … institutions [in which individuals co-operate for mutual benefit]: in co-operative relationships, it is for each individual to judge what counts as his or her own benefit’.

Finally, there is (1), the view that: ‘in a well-ordered society, co-operation for mutual benefit is a governing principle of social life’. The neo-classical account as I characterised it contains no such tenet. This element of Sugden’s account echoes John Rawls (1972, p. 84) characterisation of society ‘as a co-operative venture for mutual advantage’ in his contractarian theory of justice. Indeed, the term ‘well-ordered society’ also has a Rawlsian pedigree (see Rawls, 1972, pp. 4–5) even if Sugden may not have in mind Rawls’ precise use of this term.⁶ Can we say more than this? I think we can. Sugden cites a passage in Mill’s Utilitarianism which supports the idea that ‘given the facts of moral psychology, the right kinds of moral feelings can be sustained in a utilitarian society’ (Sugden, 2018, p. 2) and that ‘human beings have some natural propensity for “social feelings”, and that the degree to which these feelings develop depends on the structure of society’. Sugden (2018, p. 2) adds that ‘[b]y participating in relationships of mutual benefit, they form positive mental associations between benefits to themselves and benefits to others. Thus, such relationships are a nursery for the “social feelings of mankind”’. Mill’s text certainly does suggest that co-operation can be self-sustaining, and to this degree ‘psychologically stable’. He argues that people are ‘familiar with the fact of co-operating with others, and proposing to themselves a collective, not an individual interest as the aim’ so that ‘[s]o long as they are co-operating, their ends are identified with those of others’ (Mill, 1962, p. 285). While Mill makes these remarks in his defence of utilitarianism, Sugden invokes these same passages in support of the idea of ‘psychological stability’ which in the modern philosophical literature is associated with John Rawls (1972, p. 177; and Sugden, 2018, chapter 9). While Sugden may not endorse the details of Rawls’ notion of a ‘well-ordered society’, he does, on my reading, require that the principles underlying social and economic arrangements are ‘psychologically stable’. With a view to achieving such stability, his defence of the market needs to show ‘each citizen, looking forward from each point in time, that the system is likely to work to his benefit’ (Sugden, 2018, p. 202). And he adds that ‘it is hard to see how such a recommendation in defence of the market system could be given … that did not include some non-market mechanism of redistribution’. Sugden (2018, pp. 202–203) recommends a tax regime which ensures ‘partial social insurance’. Furthermore, he enlists Mill’s argument against mercantilism in the context of the possibility that redistributive taxation ‘may be an essential component of a society in which people see one another’s prosperity with goodwill’ (Sugden, 2018, p. 204). To this degree, like Mill, he believes that the ‘structure of society’ is important for co-operation to be ‘psychologically stable’.

The later chapters of The Community of Advantage – notably chapters 10 and 11 on ‘team reasoning’, ‘co-operative intentions’ and the ‘principle of mutual benefit’ – also support the first pillar of the liberal tradition as he characterises it. Furthermore Sugden (in work co-authored with Luigino Bruni; see Bruni & Sugden, 2013) provides an account of certain virtues as traits which can be derived from a view which takes the purpose of the market to be co-operation for mutual advantage. In sharp contrast with the neo-classical characterisation of the liberal tradition which I advanced on the basis of Arrow’s work, Sugden provides a defence of the market which does not ‘economise on virtue’ and none of his arguments require an assumption of self-interested motivation, which, as he also notes, has been discredited by the findings of BE (Sugden, 2018, pp. 11–12). He emphasises that ‘the market is not, or should not be, an arena of non-moral, instrumental motivation from which practices which are more genuinely social or more intrinsically valuable need to be insulated’ (Sugden, 2018, p. 281).

5. Conclusions

Robert Sugden’s The Community of Advantage advances a bold contractarian alternative to the mainstream framework of welfare economics, which ‘behavioural welfare economics’ has attempted to extend in the light of the findings of BE. In the light of a reading of J.S. Mill’s and Kenneth Arrow’s writings a sketch of the liberal tradition in neo-classical welfare economics was presented and contrasted with Sugden’s reconstruction of that tradition. I have argued that to the extent that advocates of LP attempt to provide it with a Millian pedigree in extending the neo-classical liberal tradition in the light of the findings of BE, the attempt is unsuccessful. By contrast, while Sugden’s reconstruction has a contractarian flavour and involves giving up a number of the core commitments of traditional welfare economics – each of which has a Millian pedigree – his defence of the market mechanism is nonetheless deeply rooted in some of J.S. Mill’s ideas, not merely in a few selected quotations.

Acknowledgements

A very early version of these ideas was presented at a conference on Robert Sugden’s The Community of Advantage in Rome in November 2019. I thank all those who commented there as well as Constanze Binder, Roger Crisp and especially Robert Sugden for comments. Any error is mine.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Mozaffar Qizilbash

Mozaffar Qizilbash is Honorary Professor in the Departments of Economics and Philosophy, and was previously Professor of Politics, Economics and Philosophy at the University of York. He has taught economics at the Universities of East Anglia and Southampton. His research has focussed on topics at the intersection of Economics and Philosophy, notably those relating to well-being, capability, development and moral and political theory.

Notes

1 AHP is articulated so that it is consistent with Mill’s rejection of hard paternalism. Nonetheless, the liberal tradition which follows him may be less precise than this about the form of ‘paternalism’ it rejects.

2 While Arrow makes this assumption in the context of the market, in his writings on social choice, he distinguishes people’s ‘tastes’ (which are taken to be selfish) from their ‘values’ (which might take into account ethical considerations aside from self-interest). See Arrow (1951, p. 18).

3 The same is true for other informed desire views which endorse a comparatively weak information requirement. See Qizilbash (2012 and 2021).

4 Similarly, it can be argued that LP can be manipulative and undermine autonomy (see Qizilbash, 2021) in a way that liberals would find objectionable.

5 Sugden (2018, p. 110) does, nonetheless, assume that money is always desirable.

6 For a discussion of some points on which Sugden’s contractarianism and Rawls’ original statement of his theory diverge see Qizilbash (2021).