Is the Free Market Acceptable to Everyone?

In this paper we take issue with two central claims that John Tomasi makes in Free Market Fairness (2012). The first claim is that Rawls’s difference principle can better be realized by free market institutions than it can be by state interventionist regimes such as property-owning democracy or liberal socialism. We argue that Tomasi’s narrow interpretation of the difference principle, which focuses largely on wealth and income, leaves other goods (such as control of the workplace and access to economic assets) worryingly unsatisfied. The second claim is that a wide set of economic liberties ought to be protected because they realize responsible ‘self-authorship.’ We argue that this claim also fails because, crucially, whether economic liberties serve individuals in pursuing their ambitions will depend on the nature of those ambitions and how the use of those liberties by others would affect their pursuit of them. If an expansion of liberty is good for us in some ways, but bad in others, we need to assess whether, all things considered, we would be better off with or without such expanded economic rights. We argue that the expansion Tomasi proposes is likely to fail this test.


Introduction
A commonplace defence of free markets is that they increase the autonomy of citizens over the shape of their own lives, and that this autonomy extends to all within that society, including the economically worst off. Adam Smith cites this freedom-enhancing aspect of markets as 'by far the most important of all their effects' (Smith 1981(Smith [1776, p. 412), and observes that, even the day-labourer in England enjoys a lifestyle surpassing that of the African king who stands in dominion over ten-thousand subjects (Smith 1981, p. 24). Tomasi (2012) provides a new twist to this familiar claim by extending it to Rawls's liberal egalitarian conception of justice. Free market institutions, Tomasi argues, are better equipped to realize the stated aims of Rawls's justice as fairness, than Rawls's own preferred programme of state-directed services. Tomasi adopts the Rawlsian approach to political justification that principles of political morality are sound if they are acceptable to 'free and equal citizens'. From here he argues that free market mechanisms will satisfy the claim to 'responsible self-authorship' that follows from free and equal citizenship. Furthermore, such mechanisms are capable of meeting the requirements of the difference principle with regards to economic goods, which requires any departures from equality maximally to benefit the worst off citizens in that society.
In this essay we take issue with the normative foundations of these two related claims. We argue that Tomasi's interpretation of the difference principle is too narrow-focusing as it does on the goods of wealth and income-and that he overemphasizes the importance of monetary resources. Tomasi claims that the difference principle is realized by what he calls the 'market democratic' mechanism of largely unfettered exchange. We argue that other goods covered by the difference principle properly understood (such as control of the workplace and access to economic assets) are left worryingly unsatisfied by this conception. Similarly, we argue that, the claims regarding the satisfaction of the difference principle notwithstanding, Tomasi's argument that a wide set of economic liberties ought to be protected as realizing responsible self-authorship also fails. Tomasi is too quick to conclude that extensive economic liberties are supported by responsible selfauthorship and that collective constraints on economic activity cannot benefit individuals as autonomous agents. Crucially, whether economic liberties serve individuals in pursuing their ambitions will depend on the nature of those ambitions and how the use of those liberties by others would affect their pursuit of them. This paper has the following format. The first section investigates the status of Tomasi's claim that his own conception of market democracy, free market fairness, 'realizes' socioeconomic justice specified in terms of its own distinctive interpretation of the difference principle. We show that Tomasi's defence of the free market fairness view establishes, at best, the possibility that market democratic institutions will fulfill the difference principle, but fails to address the more stringent test that it would better satisfy it than social democratic ones. The second section argues against the narrow interpretation of the difference principle as referring to income and wealth. Instead, we argue that the other goods governed by the difference principle-powers and positions of responsibility, and the social bases of self-respect-are overlooked on Tomasi's account. Contra Tomasi we show that there are good reasons for individuals to prefer such things as meaningful work and control over their working lives-as a condition of self-authorship-to increased amounts of wealth and income. Section three takes issue with Tomasi's claim that concern for citizens' self-respect does not give us reason to limit material inequalities.
Section four considers Tomasi's wider claim that thick economic rights, such as the right of individuals directly to negotiate the terms and conditions of their employment with potential employers, should be enshrined within Rawls's first principle of justice-the Basic Liberties Principle-because of the crucial role the basic liberties play in furthering individual autonomy. Following an insight from H.L.A. Hart, we argue that to know whether an expansion of liberties in the direction Tomasi suggests would be valuable, we need to know whether some people's interests might be set back by others using those liberties: for example, whether an individual's terms and conditions of employment might be worse if employers are permitted to negotiate separately with each potential employee. If an expansion of liberty is good for us in some ways, but bad in others, we need to assess whether, all things considered, we would be better off with or without such expanded economic rights. Section five illustrates our claim with regard to the issue of inheritance and opportunities for future generations. The final section of the essay considers and rejects a possible reply to our Hartian critique that Tomasi might offer, which purports to establish an intimate connection between thick economic liberties and individual self-respect.

Benefiting the Least Advantaged: Realization Versus Achievement
Tomasi argues that market democracy and, particularly, his own particular conception of it, free market fairness, 'realizes' socioeconomic justice specified in terms of its own distinctive interpretation of the difference principle. That interpretation 'seeks to maximize the personal wealth of the least well-off' (Tomasi 2012, p. 229) within the constraints of prior principles of justice, and it differs from what Tomasi calls the 'social democratic' interpretation, which includes within its account of advantage not merely personal wealth but also access to economic assets and democratic control of the workplace (Tomasi 2012, p. 228). (We criticize Tomasi's narrow interpretation of the items distributed by the difference principle in the next section.) The other key difference between market democracy and social democracy concerns the respective ways in which they seek to realize the difference principle. Whereas social democracy is portrayed as taking a direct strategy through 'statebased service programs' and 'an aggressive system of redistributive taxation' both of which limit market activity (Tomasi 2012, p. 231), market democracy seeks to benefit the least advantaged indirectly by promoting a commercial society. Sceptical of collective decision-making institutions, it seeks to expand the wealth of the least advantaged through the economic growth that, Tomasi claims, flows from relatively unregulated market exchange.
In evaluating Tomasi's argument, we need first to note the modesty of his claims. His claim is that free market fairness realizes social justice understood by reference to his particular interpretation of the difference principle, not that it achieves it. The distinction between realization and achievement relates to Tomasi's view of the enterprise of political philosophy. His aim is not to argue that market democratic institutions are better than social democratic institutions from the point of view of (his interpretation of) the difference principle. Rather, he claims that they are candidates that should be considered alongside social democratic alternatives.
Some readers might be disappointed to find that Tomasi doesn't attempt to defend the view that the difference principle supports the choice of market democratic institutions over social democratic ones. In response to that kind of disappointment, Tomasi defends a particular kind of ideal theorizing, which he thinks accords with Rawls's view. The aim of political philosophy is to describe a 'realistic utopia' in which, among other real-world facts, individuals' predictable non-compliance with principles of justice is disregarded. So the fact that particular social institutions would in the real world, here and now, produce distributions that are suboptimal with respect to the difference principle should not be taken to be a reason to reject those institutions when articulating an ideal society. In political philosophy, he says, 'the normative is appropriately constrained by the sociological possible, not the probabilistically likely' and it is, accordingly, 'insulated from practical empirical observations about the social conditions in particular societies' (Tomasi 2012, p. 213). We should not attempt to establish by reference to probabilistic judgements whether particular institutions would achieve justice, he insists. Rather, we ought to ask whether those institutions 'realize' justice-in this case, the difference principle-where realization consists in (1) having the difference principle as their aim and (2) passing a test of 'sociological realism' understood in terms of nonviolation of 'the general laws of political sociology' (Tomasi 2012, p. 219, and see pp. 215-225).
We agree with much of Tomasi's positive conception of the importance of ideal theorizing within political philosophy. That said, his claim that Rawlsians are not concerned with any kind of probability is too strong. Tomasi contrasts claims about realization, in which proposed social institutions must pass merely a modest test of non-violation of the laws of political sociology, with assessments of whether those institutions would maximize the advantage of the least advantaged in particular actual societies. This overlooks the possibility and value of a third kind of assessment: whether, given strict compliance and favourable circumstances, free market institutions would maximize the advantage of the least advantaged better, as well as, or worse than, social democratic institutions. Such an assessment would not pander to the 'complacent realism' that Tomasi and others (us included) believe we have reason to avoid; but unlike Tomasi's modest test of sociological realism, it takes into account the predictable effects of different institutions in their evaluation.
In reply to this argument for more than 'realization', Tomasi might contend that consideration of which institutions are most likely to achieve justice as fairness in a world of strict compliance and favourable circumstances is best left to political economists. Perhaps, but our sense of disappointment is unabated. Tomasi has given us an argument that it is consistent with the laws of economics that less regulation of the free market would maximize the wealth and income of the least advantaged. But that is not news to social democrats inspired by Rawls. They do not doubt that there is a possible world in which a relatively unfettered market might maximize the wealth of the least advantaged. They just don't think it likely in an ideal world of strict compliance, given the way people would predictably exercise their entitlements. Tomasi establishes only that less regulated capitalism is on the table as a possible institutional expression of the difference principle, but didn't we know that already? The real action is whether it survives scrutiny when the effects of individuals exercising the economic entitlements it affords are considered in the light of the difference principle.

How Important are Wealth and Income?
The other part of Tomasi's argument regarding the difference principle concerns the proper balance of the goods it covers and the differential levels of wealth it can be said to justify. On Tomasi's view the difference principle properly understood is concerned primarily with the maximization of wealth and income, and that this is consistent with considerable inequalities. In this section and the next we argue, contra Tomasi, that other goods such as meaningful work should be included within the index of advantage that the difference principle covers, and that the Rawlsian notion of self-respect provides grounds for limiting inequalities to which the difference principle may give rise.
The index of goods governed by the difference principle includes wealth and income, powers and positions of responsibility, and the social bases of self-respect (Tomasi 2012, p. 186). These goods are required by individuals in order to develop Rawls's two moral powers: the capacity for responsible self-authorship, and the capacity to honour their fellow citizens as responsible self-authors (Tomasi 2012, p. 184). 1 Tomasi notes that these goods may turn out to be rivalrous (Tomasi 2012, p. 186); increases in powers and positions of responsibility, say, may come at the price of reduced levels of wealth and income. One instance of this highlighted by Tomasi is the possible trade-off between higher wages-as an example of wealth and income-and the interest that individuals have, according to Rawls, in 'meaningful work in free association with others' (Tomasi 2012, p. 190;Rawls 1999, p. 257)-as an example of the powers and positions of responsibility (Tomasi 2012, p. 186).
If meaningful work is understood to include such things as greater democratic control of the workplace, with greater say over such things as the division of labour and decisions about what gets produced, then Tomasi claims overall productivity may suffer and wages may be lower than under the more hierarchical firms typical of capitalism where the emphasis is on increasing wealth (Tomasi 2012, p. 186; see also p. 189). That is, where social democracy of the Rawlsian stripe aims to realize a bundle of goods, this will mean trade-offs that lower income and wealth. Market democratic regimes, by contrast, seek primarily to maximize wealth. Is Tomasi correct? Perhaps it is possible, by contrast, to imagine a properly constructed social democracy along Rawlsian lines that could match a less regulated economy in terms of productivity and wealth generation such that workers could enjoy equally high levels of wages as they would under capitalism and find fulfillment in the workplace. 2 We might imagine a case where worker-led, stakeholding firms and enterprises with a different division of labour and managerial control that fosters enthusiasm and commitment, eliminates the wastage typically associated with hierarchical firms. Days lost to sickness by an unenthusiastic, unmotivated, and alienated workforce, or dissatisfaction at management control and surveillance of worker behavior at the micro level are common complaints about the inefficiency of capitalism. Perhaps these might be markedly reduced under social democracy with its more democratic organization of the workplace. This is an open empirical question, but its possibility is not ruled out in principle; at least, not on the grounds of probability canvassed in the previous section. Yet, it is important to note that there are reasons to think it unlikely. If such firms were really more productive, then we should wonder why, even under contemporary conditions, they are not more widespread than they actually are. If they were potentially more productive, then we should expect to see not only more of them, but entrepreneurs and investors buying up companies and turning them over to worker control in order to return a greater profit.
This empirical worry, along with Rawls's claim that citizens in a just society may be prepared to accept a lower-growth-or even stagnant-economy in return for the realization of other goods and values is sufficient to take Tomasi's claim about the rivalrous nature of such goods seriously (see Rawls 2001, p. 159). Rawls does not, it seems, rely on the idea of a social democracy outperforming laissez faire capitalism in terms of growth and productivity (see Rawls 2001, p. 178). That said, occasionally, Tomasi expresses the alternative optimistic view that much that counts as meaningful work may actually be realized under modern free market conditions (Tomasi 2012, p. 188). Current modes of production provide modern workers with considerable control over the structure and content of their roles, and increase worker bargaining power. Creativity, innovation, and independence are rewarded under modern capitalism. Gone are the days of the Fordist production line. Instead, the service economy-even the 'experience economy'-and information manipulation are centre-stage. Production is much more 'individualized' than it was even twenty years ago, and many workers experience a degree of autonomous control over their working lives that was previously unimaginable. The market demands and rewards these features, rather than stifling them.
As an empirical claim this optimism is contestable. Tomasi demonstrates something of a white-collared view of work. Whilst some roles may have changed their shape in response to market conditions in the post-industrial economy, the vast majority of people still work at jobs that are typically mundane, routinized, repetitive, and require discipline and conformity. True, those with marketable talents may enjoy significant opportunities to develop them in the way Tomasi observes, but most people likely still view work as a bind-something to be endured-and prefer to get away from it rather than spend more time at their place of employment.
It seems, then, that the conflict between goods such as income and wealth on the one hand, and meaningful work on the other, remains a legitimate concern. Faced with this potential trade-off, Tomasi suggests there are reasons to prefer a trade-off that maximizes income at the expense of these other goods (Tomasi 2012, pp. 187-189). The claim that income is preferable to meaningful work might be understood in two ways. First, it might mean, when faced with the choice, everyone would prefer greater income. Second, it might mean even if some would not prefer higher income, there are strong reasons to view higher income as more valuable than meaningful work. 3 By way of illustration of the first meaning, consider Tomasi's example of academics faced with the choice between finances for research and sitting in more departmental meetings. Framed in this way, every academic is likely to prefer more money to conduct the central aspect of their roles at the expense of any gains in democratic control of the workplace that more-or longer-meetings might bring. There are, however, reasons to doubt this claim. Even in Tomasi's narrow example, we might easily imagine academics strongly preferring the retention of professional autonomy regarding matters such as curriculum design, student admission criteria, colleague selection, and freedom of inquiry that they have traditionally enjoyed. The prospect of relinquishing control of such aspects to administrators in return for higher pay or research allowances is not self-evidently an attractive idea. It seems unlikely that preferences always pull in the direction of greater income, as Tomasi suggests. Given that we spend a considerable portion of our lives at work, individuals have a strong interest in enjoying safe, pleasant, and enjoyable working conditions and practices, and in inhabiting roles that are rewarding beyond the level of financial remuneration they bring. That some might reject greater monetary reward for workplace goods does not seem unreasonable or unlikely.
The second meaning of the claim that there are reasons to prefer increased wealth over and above meaningful work is that there are reasons independent of the preferences of individuals. To flesh out this claim we might think of money as a more flexible good than meaningful work and, as such, it enhances individual agency to a much greater extent by providing the means for the pursuit of a wide range of meaningful activities (Tomasi 2012, p. 189). 4 Again, however, we should note the strong interest we have in enjoying working (and other social) arrangements that are fulfilling. The fact that most of us will spend a significant percentage of our lives at work suggests that having control over that aspect will actively enhance our agency or autonomy in a way that more money would not. Furthermore, inhabiting roles that are mundane, routinized, and heavily controlled may actively frustrate that interest by damaging self-respect. If some occupy roles that are creative, highly autonomous and fulfilling compared with the rest who lack such employment, then this risks creating differences in social status along this dimension. This may obtain even where those who do not occupy the elite status have a level of income and wealth that is maximized under such a system (Stilz 2014, p. 434). Status is a positional good. As Rawls observes: [I]n a status system, not everyone can have the highest rank. High status assumes other positions beneath it; so if we seek a higher status for ourselves, we in effect support a scheme that entails others having lower status (Rawls 2001, p. 131).
When such attitudes are internalized by individuals, this can prevent them seeing themselves as equals with their fellow citizens. Instead, they come to see themselves (and be seen by others) as inferior. This undermines self-respect and is 'corrosive of the secure sense of standing and agency that is essential to our dignity as agents' (O'Neill 2008, p. 128). Agency is harmed when wealth and income are maximized at the expense of the other goods, not promoted.
Tomasi might be correct to point out that Rawls provides little guidance for weighing such goods other than the hope that it might be achieved once we consider our fundamental interests are in the original position (see Rawls 1999, p. 80). But, the desire for a finer-grained process that balances the various interests individuals may have does not seem likely to lead to Tomasi's conclusion that wealth and income should be maximized at the expense of the other goods. Sometimes higher income might just not be more valuable than meaningful work.

Inequality and Self-Respect
The possibility of damaging self-respect raises a wider concern with Tomasi's claims regarding the difference principle and the inequality he sees it mandating. Tomasi claims that the material requirements of self-respect are modest: 'liberal citizens can live with self-respect even in conditions of unequal wealth' (Tomasi 2012, p. 228). There are reasons to doubt this claim. Moreover, these are reasons that are internal to Rawls's construction of the difference principle. Inequalities of material conditions can fundamentally undermine self-respect. If this is the case, then reasons of self-respect can be thought to place restrictions of material inequalities.
The self-respect case against inequalities is two-fold. First, differences in economic position or differences in status from occupying certain valued positions in comparison with less valued ones might engender other inequalities or relations of domination. This might occur in areas such as social and political influence, for instance. Those who possess significant material benefits may, as a consequence, enjoy access to forms of consumption or goods that are unavailable to the less well off. They may also determine the kinds and shape of employment available, and they may exercise control over the range and availability of life plans and other activities (see Scanlon 2000, p. 44). The existence of higher levels of material resources and more socially valued roles might also spill over into the political process, allowing some to dominate or have undue influence, thus undermining the value of political equality (see Rawls 2001, pp. 245-246). Even where institutions do not formally discriminate against any class of individuals-that is, where political rights and the like are formally guaranteed-such differences can not only make representative government become such in appearance only, but also cause significant stigmatization. Those who benefit from inequalities in material resources and status may come to view those who do not as inferior because they are viewed as lacking the necessary attributes or desirable characteristics. Such status inequalities and stigmatization tends to generate the 'evils' of servility and deference in addition to the will of some to dominate others (Rawls 2001, p. 130).
Second, such inequalities are objectionable because of the sentiment or subjective feelings of inferiority they create. This directly damages an individual's self-respect. Rawls writes of the relationship between inequality and social status, that it brings us closer to what is wrong with inequality itself. Significant political and economic inequalities are often associated with inequalities of social status that encourage those of lower status to be viewed both by themselves and by others as inferior. This may arouse widespread attitudes of deference and servility on one side and a will to dominate and arrogance on the other. These effects of social and economic inequalities can be serious evils and the attitudes they engender great vices (Rawls 2001, p. 130). Rawls's notion of self-respect-taken as a person's sense of her own value, a secure conviction that her conception of the good or plan of life is worthy of pursuit, and a confidence in her ability to fulfill those intentions (Rawls 1999, p. 386)-is seriously undermined by feelings of inferiority. With regard to those who do not display the narrow range of marketable talents that are rewarded under the unfettered market, the worry is that large differences in material wealth can be objectionable because of the feelings of inferiority they create or the stigmatization as less valuable that they engender. The ways of life enjoyed by the economic elite can set norms and expectations that the worst off cannot attain, and they may well feel inferiority, even shame at not having certain excellences (Rawls 1999, p. 389; see also Scanlon 2000, p. 43). To live one's life with the thought that your abilities or accomplishments lack real worth or are inferior to those of others creates an experiential loss that is objectionable (Scanlon 2000, p. 51). Even institutions that are procedurally just may nevertheless generate the internal experience of inferiority because individuals do not enjoy 'success' as it is measured in that society. Such individuals may come to disvalue their lives, and they would do so on the basis of wrongful reasons. As such, 'the inner experience of reduced social status is associated with the loss of self-respect' (O'Neill 2008, p. 127, emphasis in original).
Even where individuals reconcile themselves to such roles, or come to actively affirm or identify with their roles, this still seems objectionable. Such identification may break the internal link with a diminution of self-respect by masking the situation and making humiliation avoidable, but such individuals would still be unjustifiably denied certain important goods, and the existence of views of inferiority and superiority that lead to domination and servility would remain. As Parfit says, an objection to material inequality can be based on the claim that, 'we may think it bad for people if they are servile or too deferential, even if this does not frustrate their desires, or affect their experienced well-being' (Parfit 2000, p. 86).
Those who benefit from significant inequalities in material conditions may also come to identify with their superior social status and see it as an important part of their own self-respect. These feelings of superiority can lead to the arrogance and willingness to dominate that Rawls mentions. In Rousseauian terminology, this is an inflamed sense of self-worth or amour propre (Rousseau 1997a(Rousseau [1754). Rousseau suggests that 'naturalizing' such status inequalities not only creates a certain arrogance that comes from falsely seeing oneself as more worthy than others, but it also creates an inner turmoil or anxiety (as well as interpersonal conflict) because the desire that others will debase themselves for our benefit can never be satisfied (see Rousseau 1997a, p. 218;Cohen 2010, p. 103; see also Stilz 2014, p. 435). Both the affluent and the poor suffer, in such cases, from defective attitudes; they base their estimations of the relative worth of their lives on falsehoods. These are great 'vices' or a serious 'evil', according to Rawls, because they violate the conception of individuals as equals upon which justice turns. In turn, this perception may undermine healthy fraternal social relations (see Rawls 1999, p. 90;Scanlon 2000, p. 51;O'Neill 2008, p. 130;Nagel 2000, p. 62). The losses to self-respect in such cases may be significant and give us strong reasons of the type Tomasi denies (Tomasi 2012, p. 230) to favour a more egalitarian distribution of wealth and income than his conception of socioeconomic justice requires.
It might be objected that if we are right in claiming that considerations of selfrespect and the avoidance of domination and servility give us reason to limit socioeconomic inequality, then the objection applies as much to Rawls's account as to Tomasi's, because on either interpretation the difference principle allows for inequalities. In reply, we simply note that although Rawls offers the empirical conjecture that his two principles are sufficient for self-respect, he is open to the suggestion that they are not and, in particular, that the inequalities permitted by the difference principle may need to be reined in to protect the self-respect of the least advantaged (Rawls 1999, pp. 478-479).

Thick Economic Liberty, Self-Authorship, and Hart's Insight
Even if free market institutions fail to satisfy the difference principle correctly understood, Tomasi has an alternative defence of them, namely, that they ought to be protected under a suitably revised basic liberty principle that takes priority over considerations of socioeconomic justice. Following Rawls, he insists that 'free and equal citizens' have a claim to institutions that facilitate their 'responsible selfauthorship', which, as Rawls understands it, is the interest in developing, deliberating about, and pursuing a conception of what it means to live well (Rawls 2001, p. 19). Tomasi's argument is that, properly understood, that interest supports capitalist economic freedoms in addition to the list of basic liberties that Rawls defends.
The key distinction that Tomasi draws is between 'thin' and 'thick' conceptions of economic liberty (see Tomasi 2012, pp. 68-69). He attributes to Rawls the thin conception, which asserts that individuals have a limited right to occupational choice and a right to personal property. By contrast Tomasi follows Nickel in setting out a thicker conception of economic liberty that includes the following elements: others are not permitted by justice to force individuals into particular jobs or to prevent them from taking jobs offered by others; individuals should have the legal right to negotiate the terms of conditions of their employment; they should not be prevented by others from establishing, running or closing their own business ventures; and their use of their material resources-savings and investments, exchanges with and gifts to others-should not be prohibited or prevented by others and should not be taxed for redistributive purposes (Tomasi 2012, pp. 22-23;citing Nickel 2000).
Tomasi's strategy for defending thick economic liberties is to show how responsible self-authorship is protected by these liberties, but impeded by merely thin economic liberties. He agrees with Rawls that occupational choice-understood as not having one's job set by anyone else-is a requirement of individuals pursuing their own views of the good life. After all, one's working life occupies a good proportion of one's lifetime and, therefore, working in an occupation set by another can hinder one in pursuing one's own projects. Moreover, one's work might have intrinsic or expressive value: for some, living well is constituted by certain kinds of employment; others value work that enables them to express their convictions about what is important-to help the needy or to educate children, for example.
Tomasi argues that the interests that are served by freedom of occupational choice extend to defend thicker economic liberties, including the right of individuals to negotiate for themselves where they work, how many hours they work, in short, the terms and conditions of their employment. The absence of these further liberties, he writes, 'would truncate the ability of those people to be responsible authors of their own lives' (Tomasi 2012, p. 77). Employment contracts would then be decided by others rather than ourselves. He makes similar claims about other thick economic liberties.
Is Tomasi right? It is important first to note that his self-authorship argument establishes only a ceteris paribus claim, which asserts that other things being equal expanding an individual's economic liberty, by giving her the power to choose the terms and conditions of her employment, enhances her ability to pursue her own conception of the good. But further claims need defending for such a ceteris paribus claim to become an all-things-considered claim about the desirability of a societywide expansion of liberty. 5 This is one of Hart's insights in his important critical review of Rawls's liberty principle: Any scheme providing for the general distribution in society of liberty of action necessarily does two things: first, it confers on individuals the advantage of that liberty, but secondly, it exposes them to whatever disadvantages the practices of that liberty by others may entail for them… So whether or not it is in any man's interest to choose that any specific liberty should be generally distributed depends on whether the advantages for him of the exercise of that liberty outweigh the various disadvantages for him of its general practice by others (Hart 1989, pp. 247-248).
Following Hart, before we accept that what appears to be an improvement with respect to pursuing one's conception of the good really is an improvement for an individual, we need to know (a) whether others' use of that same liberty has negative consequences for her, and (b) whether, taking into account those negative consequences, she is better off with respect to pursuing her conception of the good with or without the expanded liberty to negotiate her terms and conditions of employment.
To operationalize Hart's test consider two stylized cases.
Individualized negotiation In this regime, each individual negotiates directly with her employer how many hours she works, the amount of vacation she takes, her wage or salary for the job, and the basic terms and conditions of her employment.
Collectively constrained negotiation In this regime, employment contracts are constrained by rules agreed in negotiation with trade unions or the government: a limit on the working day or week, a minimum number of days of vacation, salary arrangements that require everyone who performs the same job to have the same wage, and further terms and conditions that protect the employee from various kinds of harm or disrespect. 6 The question that confronts us is whether the freedom individually to negotiate the terms and conditions of work with an employer protects an individual in pursuing her conception of the good. The answer is 'not obviously'. Like Odysseus who by binding himself to his ship's mast successfully pursued his goal of hearing the beautiful Siren voices without being killed on the rocks, it might be that many individuals pursue their employment-related aims better if constraints are in place that prevent them from being burdened by their lack of individual bargaining power.
One feature of individualized negotiation that is not emphasized by Tomasi is that it augments the power of employers who become free to consider what different would-be employees have to lose by being denied the job in question and to tailor employment offers accordingly. Those individuals who have more to lose tend to Footnote 5 continued question remains relevant even if it is accepted that the reasons to protect the liberties do not always defeat other reasons of political morality. 6 To clarify, collectively constrained negotiation does not prohibit variations in employment contracts within the parameters set by the collectively set rules; but, unlike individualized negotiation, it does prohibit employers from offering worse terms and conditions than those set by the collectively determined rules. accept worse terms and conditions and a lower salary than those who have less to lose. That is good for the employer, but worse for some employees than collectively constrained negotiation in which employees are protected from having their worse bargaining position reflected in their employment contract. To the extent that we care about individuals having the opportunity to pursue their conceptions of the good we must attend not merely to how the law constrains them, but how their pursuit of their goals might be impeded if employers are free to exploit information about individuals' different bargaining positions. This is a familiar argument presented by economists and philosophers going back at least to Rousseau (1997b), but one that Tomasi does not systematically consider.
Accordingly, we need to know more about the effects of individualized negotiation on the terms and conditions of employment that are offered before we embrace it. We can accept that it is of the utmost importance that individuals devise, reflect on, and pursue their own ethical goals. Since many people have goals with respect to employment that include secure employment and decent terms and conditions of work, it is an open empirical question whether their opportunity to pursue that goal is served by individualized or collectively constrained negotiation.
If an individual were able to pursue her employment-related ambitions less well via individualized negotiation, and objected to such a regime for that reason, it would seem that her objection to that economic liberty should be taken seriously. Of course, others might object to collectively constrained negotiation. In the face of this disagreement we need a fair way of choosing between these two regimes for determining citizens' terms and conditions of work. There might be various ways of selecting a fair policy. One way would be to assess whether, given what they value, individuals would prefer individualized or collectively constrained negotiation if they knew the society-wide distribution of bargaining power in the economy but were ignorant of whether they themselves are on the up-or down-side of that distribution. 7 Perhaps there are other ways of resolving the conflicting claims. Our modest observation is that the appeal to the good of individuals having the opportunity to pursue their conceptions of the good does not generate the conclusion that this economic liberty is in everyone's interest.
Tomasi might offer an argument that seeks to side-step empirical questions about the effects of different negotiation regimes on people's opportunities to pursue their goals. The argument is that individualized negotiation uniquely protects selfauthorship, because being the author of her own life requires the individual to make her own occupational choices rather than have them constrained by others, even if the effects of occupational choice are worse for her than if she were constrained by collective agreements. The task of living well, we might say, falls on the individual and it is therefore her right to make choices that carry the possibility of her living poorly as well as successfully. So, it is for individuals to choose how to work; that is their right even if the effect of individualized negotiation delivers worse employment prospects for some or most. Personal autonomy requires us to be the agent of our lives. We have reasons to do certain things, rather than to have items given to us by others-we should choose between the alternatives and take steps to realize the goals we decide are valuable. Even if it turns out that individuals exercising their agency generates for them less enjoyment, security, welfare, or whatever they value (or is valuable) our self-authorship demands that others not usurp their right to decide for themselves how their lives should go. Collectively constrained negotiation is incompatible with individuals exercising their agency over the central aspects of their lives, even if it benefits them. It is an instance of wrongful paternalism that has no place within a conception of political morality that protects freedom.
But notice that there are two ways in which collectively constrained negotiation might benefit people. First, it might constrain employers in ways that deliver for individuals more of various goods, like contentment or a sense of security, that make no reference to the exercise of agency. Second, it might constrain them in ways that delivers to individuals agency-related opportunities: opportunities for individuals to choose how to live their lives either in work, by having a voice in the way their employment is organized, or outside work, by having more leisure time or income that enables them to choose between a wider set of options. Tomasi is right to emphasise that a concern for autonomy prioritizes choice and the exercise of agency, which may rule out the first kind of constraint, but wrong to think that it follows that individuals should never be constrained. An individual whose choice with respect to employment is limited by the government is deprived of the opportunity to decide what to do in the face of that choice, but the constraint might widen her opportunity to exercise her choice in other parts of her life and she might prefer being so constrained just for that reason.
If our argument is right, then Tomasi is too quick to conclude that responsible self-authorship supports individualized negotiation of employment contracts, and too quick to claim that collectively constrained employment contracts serve individuals merely as recipients of goods but not as autonomous agents. Whether thick economic liberties serve an individual in pursuing her ambitions depends on the nature of those ambitions and how others' use of those liberties would affect her pursuit of them.

Self-Authorship, Parents and Children
To recognize the more general relevance of Hart's insight and explore a different feature of Tomasi's view consider a different economic liberty: the freedom of individuals to give and others to receive without interference from others in the form of redistributive taxation. The taxation of estates fails to protect selfauthorship, some argue, because it prevents individuals from expressing and pursuing one of their deepest ambitions-their commitment to their children.
To judge this claim we need, once again, to run Hart's test and ask whether the liberty to give and to receive is, all things considered, in everyone's interest, taking into account both (a) the value to an individual of her being free to attend to her children's interests via gifts or posthumous bequests and (b) the effects of other individuals attending to the interests of their children in the same way. To fix ideas, let us consider two regimes that might govern posthumous bequests to children: Unrestricted bequest individuals are legally free posthumously to bequeath to others as they choose; such bequests are not taxed. Restricted bequest posthumous bequests are taxed and the tax receipts are shared between citizens.
Parents might think that the opportunity to promote their children's financial interests by posthumous bequest is valuable for several reasons, but two seem prominent. One simple reason is the well-being of their children. We might value the opportunity to give to our children because it is important to us that their lives go better than they otherwise would, and, other things equal, having more resources enables them to live better lives. The second putative reason is that it is valuable to give to one's children regardless of whether the system of giving makes one's child better or worse off than she would be compared to a different system, such as one in which inherited resources are distributed to the next generation equally or in ways that benefit the least advantaged.
To the extent that one is motivated by a concern for one's child's well-being some will favour unrestricted bequest, some will favour restricted bequest. If an individual is beset by the bad luck of having a disability that prevents her from securing a well-paid occupation it might be that her child will secure more resources from restricted bequest. Of course, whether or not that is the case depends upon several factors. Some argue that taxing bequests is a disincentive to work. If that were true, it might be that, although unrestricted bequest fails to share the wealth generated by the previous generation, the child who loses out with respect to inheritance would gain in other ways, by living in an economy that produces more than it would under restricted bequest. Others have argued that unrestricted bequest is economically better because it encourages savings that can be used to invest in companies that generate economic growth. Again, if true, that would count in favour of abolishing taxes on estates (McCaffery 1994).
But the incentive and savings arguments for unrestricted bequest need more work, because it is not clear that individuals' attitudes to work or to savings are significantly affected by marginal changes to taxes on estates (Arrondel and Masson 2012). In any event, there may be an optimal tax on estates that, all things considered, generates most for the children of low earners, which, in all likelihood, is positive. To the extent that parents are concerned about their children's wellbeing, restricted bequest seems the rational choice for many, perhaps most or, at least, those who lack marketable skills.
Of course, this argument for restricted bequest is threatened by the second reason individuals might have to give to their children. Individuals might regard it as important that they give to their children, even if their children would benefit more by having a share of everyone's posthumous wealth. That ambition makes sense in many cases. It matters to individuals that they can bequeath the medals and memorabilia they obtained in wartime, or their wedding rings, and that remains the case even when the medals are merely service medals rather than the Distinguished Flying Cross, or the wedding ring made from nickel rather than platinum. They know, let's suppose, that their child would be better off having a share of all the previous generation's Is the Free Market Acceptable to Everyone? medals and wedding rings, but they insist nonetheless on the importance of having the opportunity to bequeath those items that have significance for them.
But although we can make sense of the ambition to give even when it is known that one's child would do worse than she would do if inheritance were shared, we need to weigh that ambition in the balance with one's concern for one's child, which may favour restricted bequest. Of course, there might be some mixed regime in which certain treasured personal possessions, such as medals or rings, may be bequeathed without tax liability, but other items such as houses or businesses are liable to be taxed. Our point is merely that when our reasons to bequeath goods to our children are examined more carefully, one very weighty reason that motivates us is that we want our children to do well and, to the extent that that reason is in play, the less productive may have a reason to favour restricted bequest because their children fare better under that system than they do under unrestricted bequest.
As before, when people's values are examined in detail and considered in light of information about the predictable effects of others' use of the proposed liberty as well as their own, those values might not be realized best by having more freedom to bequeath. They might be better realized in an economic regime that places certain limits on people's freedom to transfer their property to their own particular children via posthumous bequests. Accordingly, a simple reply to the unrestricted bequest proposal that might be made by a parent who has few marketable assets is that the value that weighs most heavily with her is her child's well being. If her child would fare better under a different (restricted or sharing) regime, it seems that unrestricted bequest is not justifiable to her given the values that animate her life. 8 We have been discussing posthumous bequests by assuming that our policy should reflect the values and ambitions that move individuals and how they play out in different circumstances. If that approach is broadly right it might be developed at greater length to attempt to identify what would be a fair policy given the fact that people hold somewhat different views about the nature and weight of their reasons to give to their children. And it must incorporate the fact that the interests of children have standing in their own right, which need somehow to be factored into the equation. 9 8 Others have criticized Tomasi's treatment of gifts and bequests on different grounds. Carens (2014, pp. 284-286) takes issue with Tomasi's permissive conception of bequests, but he does so by reference to the entitlements of children. Our critique of Tomasi focuses largely on the ambitions of parents. Stilz (2014, pp. 430-436) appeals to the ideal of fair equality of opportunity and concerns about domination, which are overlooked by Tomasi. In response, Tomasi (2014, pp. 457-461) has revised his account accepting the need for a political community to avoid 'class domination' and proposing a progressive tax on bequests or inheritances to protect against that outcome. Our critique of Tomasi's account of transfers between generations rests on the more modest claim that it is insensitive to the ethical ambitions people have with regard to their children. 9 For an attempt to identify an inheritance regime that is responsive to people's ambitions in a fair manner, see Clayton (2012). That account is heavily indebted to Dworkin's account of liberal equality. It is notable that Tomasi does not seriously engage with Dworkin's conception of political morality (Dworkin 2000(Dworkin , 2011. That is unfortunate, because Dworkin argues for an egalitarian view of justice in which the market plays an essential, if constrained, role. Among those whom Tomasi describes as 'high' liberals, Dworkin is the theorist who takes economic liberties most seriously; but, nonetheless, Dworkin argues for redistributive arrangements that are typically associated with social democracy. For a nice But Tomasi might insist that we have misread his appeal to self-authorship. He might say that it doesn't matter whether inheritance arrangements are responsive in a fair manner to people's ambitions or values. He might say that it is simply bad for people to be recipients rather than agents. People's lives go worse if they do not face challenges that require them to exercise skill and judgement to succeed and which come with the possibility of failure. He appears to court such a view when he notes admiringly Murray's claim that social democratic arrangements that limit the risks people are exposed to deprives them of the opportunity to meet their (and, more chillingly, their children's) needs themselves and to face up to the challenge of living well (Tomasi 2012, pp. 80, 182;citing Murray 2009).
Perhaps, then, we ought to take Tomasi's remarks about the need for opportunities to ensure that one's children are well provided for through one's own efforts, rather than by sharing the assets produced by others, to be claims about what is good for people. If so, he may have a basis for replying to Hart's insight: it might be true that an individual's aim of ensuring that her child is well provided for is satisfied less well, but unrestricted bequest gives her something more important, namely, the opportunity to provide for her child by her own efforts.
If that is Tomasi's claim, however, there is a worry about whether his view is consistent with the democratic justificatory requirement he takes from Rawls, that principles of justice and their justification must be acceptable to free and equal citizens who disagree about what it means to live successful lives. True, some citizens have ethical ambitions that involve the exercise of agency and of overcoming challenges themselves. But a citizen might reasonably reject the view that her life goes worse if her children receive resources via shared inheritance rather than because she has, by her own efforts, succeeded in employment and transferred resources to them directly. Accordingly, it is difficult to see how a regime of unrestricted bequest can be justified to everyone given that fact.

Self-Respect
Tomasi's reply to this kind of objection is to appeal to Rawls's primary good of the social bases of self-respect. Everyone has an interest, regardless of her specific conception of the good, in living under social and political arrangements that 'are normally essential if citizens are to have a lively sense of their worth as persons and to be able to advance their ends with self-confidence' (Rawls 2001, p. 59). If Tomasi can establish that free market institutions are indeed essential for individual selfrespect, then he can argue that they are good for individuals in a way that does not violate the justificatory constraint of acceptability to reasonable people who disagree about what it means to live well. 10 Footnote 9 continued review of late Twentieth Century debates about distributive justice and Dworkin's contribution to them, which departs from Tomasi's portrayal of the recent history of political philosophy, see Williams (2006).
To establish the connection, Tomasi proceeds in two steps. First, he argues that self-respect, understood as a sense of one's own worth and one's ability to pursue one's ends with self-confidence, is 'diminished if one is not (and so cannot think of oneself as) the central cause of the life one is leading' (Tomasi 2012, p. 83). Second, he claims that thick economic liberties afford opportunities for individuals to be the central causes of their lives; social democratic arrangements, which deny individuals the right to decide for themselves how their working lives should go and the opportunity to have their choices and actions appraised by others, are (the strong claim) inconsistent with self-respect so understood, or (the weaker claim) threaten it.
It is difficult to evaluate Tomasi's argument because he does not provide the reader with the details of what he means by an individual being the central cause of her own life. Nevertheless, on the basis of our rough understanding of it, let us consider his argument by reference to the two economic liberties that have been our focus-unrestricted bequest and individualized negotiation. With respect to bequests, it is unclear why being the central cause of one's own life supports the economic liberty of parents to make posthumous bequests or gifts inter vivos to their children. If anything, heavily constraining bequests would seem to be supported by our concern for the self-respect of children of wealthy parents who would be deprived of the opportunity to see themselves as causes of their own success if their parents 'provide them with all the ''gilded material means'' in the world' (Tomasi 2012, p. 83). 11 It seems that the economic liberty of giving and receiving should be severely curtailed if self-respect requires individuals to pull themselves up by their own bootstraps, as Tomasi seems to suggest. If, on the other hand, it is accepted that self-respect is not threatened by individuals receiving resources from others, then self-respect cannot do the required work in blocking an argument for restricted bequest in which estates are taxed and the proceeds shared between everyone.
What about the economic right to negotiate the terms and conditions of one's own employment? Does that find support in the idea that individuals should be the central causes of their own lives? Tomasi thinks that it does, because laws prescribing particular terms and conditions of employment in the form of minimum wage legislation, for example, straightforwardly deny individuals' the opportunity to decide these matters for themselves (Tomasi 2012, pp. 83-84). However, before we accept Tomasi's conclusion, we need to note that a free market generates unequal opportunities with respect to employment: while individuals who happen to be blessed with multiple marketable talents might have very many employment opportunities between which they can choose, those who lack marketable talents have a more restricted choice. Self-respect, understood as doing well as a consequence of one's own activities (Tomasi 2012, p. 83), does not appear to be threatened by the fact that one's employment opportunities are wide ranging. A person who is lucky in her biological endowment can still make something of her life and can be appraised by others according to how well she succeeds in performing, even when the opportunities that she faces are great. If that is the case, however, it is unclear why self-respect is inconsistent with protecting an 11 See Carens (2014, p. 285) who also makes this observation.
individual's employment opportunities through legislation that limits employers' ability to respond to low-skilled workers' poor bargaining position by offering them employment opportunities that are unattractive. True, as we noted before, the individual's opportunity to decide between certain options-in some cases between unattractive employment and poverty-is absent; but it remains the case that she has to choose between different options, and she can be appraised for choosing more or less well and for performing better or worse in the options she adopts.
If we are right, then the appeal to self-respect cannot do the work Tomasi thinks it can in supporting thick economic liberties. If our earlier arguments are also sound, then Tomasi faces a dilemma. He might develop different arguments that show that protective employment legislation is bad for individuals according to some detailed conception of well-being. However, such arguments risk political perfectionism, which is inconsistent with the liberal justificatory framework within which he aspires to work. Alternatively, he might try to defend thick economic liberties through an appeal to what everyone cares about, right or wrong. But, as we have suggested by developing Hart's insight, such an argument-for at least two key economic liberties-is unlikely to succeed, because the economic liberties are valued by some but not others. If our theory of distributive justice is to be sensitive to the ambitions individuals have for their own lives, then we need to identify a fair way of resolving such conflicting interests.