Public justification,gender,and the family

Abstract

Social norms regulating carework and social reproduction tend to be inegalitarian. At the same time, such norms often play a crucial role when we plan our lives. How can we criticise objectionable practices while ensuring that people can organise their lives around meaningful and predictable rules? Gerald Gaus argues that only ‘publicly justified’ rules, rules that everyone would prefer over ‘blameless liberty,’ should be followed. In this paper, we uncover the inegalitarian implications of this feature of Gaus's framework. We show that because a society without clear social norms for how social reproduction and care work ought to be organised would be so unattractive, inegalitarian rules would pass Gaus's test. They would pass this test since they would nevertheless be better than ‘blameless liberty.’ Those who are disproportionately burdened by a rule are faced with the daunting task of showing that they would be better off under no rule, instead of merely having to show that they would be better off with a different rule.

Rigid, the skeleton of habit alone upholds the human frame.

Virginia Woolf, Mrs Dalloway.

A significant part of our social world consists of inegalitarian, exploitative, and unjust practices. Some of these practices ‘merely’ place disproportionate burdens on particular groups, whereas others are seriously harmful to them. At the same time, people plan and organise their lives according to established practices and the expectation that these will persist. They believe, to some extent, implicitly or explicitly, that they will be able to continue living as they always have. This is especially true when it comes to how systems for governing intimate relationships and social reproduction are set up. Many family practices are gravely inegalitarian. Yet, the norms and practices of the family are central to our identities and our ways of life. Balancing these considerations is therefore an essential task. How can we criticise and reform objectionable practices while ensuring that people can organise their lives around meaningful and predictable rules?

Gerald Gaus has recently provided an impressive answer to this question. Gaus (2010) argues that we are only morally obliged to follow and enforce publicly justified rules, be they moral or political. Such rules are those that are acceptable to everyone who is subject to their authority. Rules that some people reject do not have the appropriate moral force. On the face of it, this seems to take care of the worry about inegalitarian practices—surely at least some people would reject them, so they could not have the appropriate moral status. Further, Gaus maintains that when a publicly justified rule is widely followed, morality demands that we continue to follow it: we cannot follow our own preferred rule instead. Taken together, this appears to imply that once we reach a state of affairs where all rules governing the family are publicly justified, everyone will have reason to accept these rules, and they will be able to plan their lives accordingly. Thus, the tension looks resolved.

Gaus's critics have argued that the set of publicly justified rules would be too small, and therefore, too restrictive. James Boettcher (2015) objects to the apparent ‘classical liberal tilt’ of the framework, Andrew Lister (2010, 2013) suggests that decisions would always be whatever the most libertarian citizen decides, Christopher Eberle (2011) argues that it will leave the state unable to deliver many important goods, and David Enoch (2015) worries that the only possible outcome is anarchy—no rule can be publicly justified in the sense that Gaus is concerned with (see also Schultz-Bergin, 2021).

In what follows, we pursue the opposite line of critique. We argue that the framework is too permissive. Hence, the problem as we see it is not that only a limited set of rules will be publicly justified. Rather, when non-legal norms are evaluated through Gaus's framework it will lead to the conclusion that even objectionably inegalitarian rules can have the appropriate moral status. ¹ In this way, the tension is not resolved—Gaus's theory unduly privileges the status quo.

In short, our argument uncovers the implications of a central feature of what it takes for a rule to be publicly justified on Gaus's view: that everyone should agree that it is better to have it in place over having no rule at all. Using inegalitarian family norms as our leading example, we show that the problem is that because a society without norms for how social reproduction and care work ought to be organised would be so unattractive, inegalitarian rules would pass Gaus's test. They would pass this test because they would nevertheless be better than ‘blameless liberty.’ Those who are disproportionately burdened by a rule are faced with the daunting task of showing that they would be better off under no rule, instead of merely having to show that they would be better off with a different rule. ²

We begin by setting out the central features of Gaus's theory of morality and public justification. After this, we show that established inegalitarian family norms would be publicly justified on Gaus's account. We then demonstrate that neither the formal constraints that Gaus incorporates in his theory can solve the problem, nor can the fact that the rules have a patriarchal origin. We finally consider whether Gaus's framework nevertheless is the best we can hope for. ³

The set of socially optimal eligible rules

The motivating thought behind Gaus's theory of public justification is that people need to have sufficient reason to accept any rule that binds them. This includes moral rules, social norms and political rules, like laws. Gaus argues that binding rules without this feature fail to treat people as free equals. ⁴ A rule is publicly justified if everyone, after moderate idealisation which removes obvious epistemic errors and flaws, would prefer having the rule in place over not having any rule in the particular domain. ⁵

How do we determine which rules are publicly justified? First, Gaus says, we should limit ourselves to those rules that are within the so-called socially optimal eligible set of moral rules. To understand what the socially optimal eligible set is, we can begin by picturing all possible rules for a given domain. We then remove those rules that are ineligible because some ‘Members of the Public’—that is, some of the moderately idealised citizens—do not have reason to endorse them over having no moral authority at all. In other words, if a Member of the Public would prefer having no rule at all in place for regulating the domain over having the rule, it is removed, because this also makes it socially ineligible. Simply put, the reason that these rules are not within the socially optimal eligible set is because ‘at least one free and equal person does not have reason to endorse the authority of its claims’ (Gaus, 2010: 322).

After this, we end up with a set of rules that are eligible. Some of these rules, however, might be Pareto dominated. If there is a rule y, and everyone prefers having rule x in place instead of it, y cannot be a part of the socially optimal eligible set, because acting on y over x would manifest collective irrationality. Hence, a rule is in the socially optimal eligible set of rules if it is neither ‘ineligible in anyone's ranking, nor… dominated by any other member of the set.’ (Boettcher, 2015: 196; Gaus, 2010: 322–323).

Now, let us suppose that we want to determine whether people have genuinely moral reasons for following a ‘widely observed’ social rule. We want to know if it generates moral duties, and if people can legitimately hold each other accountable to it. In order to know this, we must, according to Gaus (2010: 425), ‘test whether it is acceptable from the perspective of free and equal moral persons.’ And what is acceptable, in this sense, is that which is within the optimal eligible set. So, if a rule x is in the optimal eligible set,

then x as a current social rule is now the basis of a moral equilibrium: a rule that has been converged upon and can be freely followed, and whose authoritative nature can be acknowledged by each… (Gaus, 2010: 425)

Importantly, on Gaus’s view, the fact that we have converged on a rule from the socially optimal eligible set provides conclusive moral reasons for following this particular rule over alternative, not-widely-followed proposals that are also part of the optimal eligible set. Hence, if ‘an existing rule is within the optimal eligible set, it is publicly justified for that reason—just because it is the existing rule’ (Gaus, 2010: 425). Publicly justified rules, then, are those widely observed rules that are within the socially optimal eligible set.

As one can see, for Gaus, arriving at a set of publicly justified abstract principles does not get us far enough. We need rules ‘that are sufficiently fine-grained to serve as the basis of our actual social life’ (Gaus, 2010: 42). Gaus (2010: 414) wants his theory to be able to both draw on, and yield determinate conclusions for, the real world. The crucial move employed for doing so is appealing to coordination on social rules and practices in real societies. To show that coordination among people who disagree is possible, Gaus constructs what he calls the ‘Kantian Coordination Game’ where the players evaluate possible rules against ‘blameless liberty’. Blameless liberty, for Gaus, is the default. It means that no agent has standing or authority over any other agent to ask them to perform (or not perform) an action. Under blameless liberty, Betty cannot demand that Alf help her do something that he does not want to do. This stands in contrast to if there is a moral rule that says that people in situations like Alf's ought to help people in situations like Betty's. If such a rule exists, Betty has the standing to demand that Alf complies (Gaus, 2010: 319–20).

What would individuals prefer over blameless liberty? One thing, Gaus argues, is that they would all agree to a set of liberal rights, such as rights not to be coerced, a right to freedom of thought, and so on. And there is a presumption in favour of liberty—that is, Betty asking Alf to do something will always require justification, but Alf refusing to help does not require justification unless a moral rule has been established (Gaus, 2010: ch 6). Moreover, Gaus believes that Members of the Public would also agree to a fairly determinate set of social rules. Individuals will have reason to coordinate on an equilibrium—otherwise they cannot reap the benefits of social cooperation. And once ‘society has chosen a rule, if the rule in equilibrium is also a member of the optimal eligible set, we have created through our actual interdependent choices what impartial reason could not deliver: a uniquely justified rule.’ (Gaus, 2010: 402)

In the political realm, it is widely accepted that Gaus's justificatory liberalism has a classically liberal tilt. Since the framework presupposes a ‘default of simply not having any coercive law or policy at all’ (Boettcher, 2019: 195), there is a higher burden of proof for implementing a rule than for retaining a situation with no rule in place. Hence, if Betty proposes a rule saying that everyone should get one month of paid parental leave, it is enough for Alf to say that he would rather not have any laws or rules regulating family leave, because he prefers blameless liberty. This entails that the proposal cannot be publicly justified, and so the state would lack the legitimate authority to enact family leave provisions.

Naturally, many egalitarians would object to this since they worry that it will leave the state unable to pursue many of their favoured policy goals. And, indeed, we believe that there are good reasons for rejecting a framework that leaves the state without the power to enact policies that we take to be essential for realising liberal egalitarian justice. However, the worry that too few measures are justified in the political domain is not our focus. Instead, we argue here that the framework (also) suffers from the opposite problem in the interpersonal domain. It allows too many rules, even those that are objectionably inegalitarian. We turn to systems for social reproduction to show this.

Injustice in the family

All human societies need some way of tending to the young, the elderly, and the fragile. If we grant Gaus's assumption that under blameless liberty, Alf has no moral duty to assist Betty, or anyone else, it seems that, in order for this to work, some rules must be in place to govern the activity of care work. Without predictable rules, we cannot predictably determine that when someone gets pregnant, there are structures of support in place for them, or that when a new child is born, someone is there to care for it. Similar things can be said for when people fall ill, need rest, or grow old.

Many societies (existing and historical) organise care work and intimate relationships in a way that asks women to shoulder greater burdens than men. When family members fall ill, social norms often dictate that mothers and daughters have a greater responsibility to care for them than have fathers and sons. Women are also tasked with raising children, caring for the house, and so on, to a greater extent than men. These are inegalitarian family norms. It should be uncontroversial to point out that such norms are widely adhered to in many places. As Wendy Brown (1992: 7–34, 25) puts it, ‘the problem here is one most feminists could recite in their sleep’. Societies ‘like ours’ are plagued by a long history of a gendered division of labour and of excluding women. And, as Okin (2004: 1550; see also MacKinnon, 1987) says, in this history

all major social institutions have been constructed on the assumption that someone other than their participants is taking care of the young, the old, the sick, and even the participants’ own daily needs.

We assume that it is unjust and morally wrong to make such assumptions and to organise societies in this way; ⁶ as bell hooks (2000: 50) argues, ‘when women in the home spend all their time attending to the needs of others, home is a workplace for her, not a site of relaxation, comfort, and pleasure.’ The question that will concern us in the remainder of this paper is this: could Gaus's view support this conclusion?

Gaus's theory asks us to consider each norm or social rule individually, holding everything else fixed. We should assume that other rules are in place, and ask whether we would prefer the rule that we test over not having any rule at all for governing that particular domain. When we perform this test, some inegalitarian rules might be excluded on the basis that they violate basic claims of agency. For instance, if Gaus is right that everyone would agree on basic rights, perhaps gravely unjust and wrongful practices like marital rape cannot be publicly justified insofar as they violate basic rights. ⁷ If so, we have moral grounds for criticising rules that allow marital rape. This is a small victory, however, because a great deal of inegalitarian family norms are not violations of basic rights in this direct way.

Could inegalitarian family norms that do not amount to rights-violations be publicly justified? To help us in thinking this through, let us picture Alf and Betty, a heterosexual couple expecting a baby. Raising a child is costly, and children need a lot of time with adults who care for them. So Alf and Betty will need some way of allocating the time between generating an income and caring for the child. How should their time be spent? What rule should they act upon in deciding who should do what?

One suggestion is that Alf and Betty should follow whatever social norm or rule their society has in place for governing how time should be allocated. Of course, we could imagine people who operate freely of such rules, who simply negotiate individually about who should do what. The question that Gaus's framework is concerned with, however, is the general rule that applies to people in Alf and Betty's situation, not whether Alf and Betty could opt out from the rule if both of them agree to do so. The relevant consideration is that if there is a rule, we cannot unilaterally defect from it. In Gaus's framework, people have moral standing to hold each other accountable to publicly justified rules, meaning that if Betty and Alf fail to come to an agreement on their own terms, they should follow the rules laid out for them by society. ⁸

In principle, there are many different candidate rules that Alf and Betty's society could organise care work around: each parent could be asked to take an equal share of the care work and of the paid labour, the care could be organised such that additional carers who are not the biological parents assist with it, the care could be mainly left to state institutions while the parents bring in money, and so on. However, one common way of organising things is, as mentioned above, the gendered division of labour. More formally, one might conceptualise the gendered division of labour as a social rule that says that in heterosexual relationships, women should carry the main responsibility for house- and care work, whereas men have a greater responsibility to provide financially for the family (although of course, most women also work to provide financial gains for the family, alongside their carework responsibilities, ⁹ and most fathers also provide some care). In that sense, the rule says both that the woman should be the primary carer for the child (she should take time off work if the child gets sick, she should work less than the man (or not at all), she should keep track of the things that regularly need to be taken care of in the caring for the child) and that the man should make sure that the family is financially secure (he should work longer hours, strive for promotions, and so on).

Let us imagine that in the society where Alf and Betty live, the gendered division of labour rule is widely adhered to: the expectation among expecting couples in this society is that the gendered division of labour rule should be followed. Most women have the primary responsibility for the carework, and most men are assigned the role of the ‘ideal worker’ who can ‘delegate virtually all domestic responsibilities’ (Williams, 2000: 91n25; Williams, 2002: 31, see also Edlund and Öun, 2016). ¹⁰ If the rule would be publicly justified, it would on Gaus's view entail that a woman who wanted to organise the care for her child differently would not have the moral standing to demand of her partner that the time between income generation and care be divided equally. And, vice versa, a man who wanted to spend more time with his children would come up against expectations that he instead takes the main share of the financial responsibilities. ¹¹

Let us again quickly run through the steps that are required to test whether a rule is publicly justified. First, we ask whether every Member of the Public prefers the gendered division of labour rule to a situation where no rules govern how children are cared for. Then, we examine if the rule is Pareto dominated, that is, if everyone would prefer another rule to it. It seems to us straightforwardly true that the Pareto criterion could not exclude most inegalitarian rules, including the gendered division of labour rule. In an inegalitarian society, there will be people who believe that the best way of organising social reproduction is by keeping things as they are (or at least they do not all agree that an alternative is better). The currently existing practices will be the top choices of some rankings. If so, the inegalitarian family norms cannot be excluded from the socially optimal eligible set on Pareto grounds.

The crucial question, thus, is if each moderately idealised citizen would prefer having the gendered division of labour rule over having no rule at all in this domain or if this rule would be rejected. ¹² Perhaps it would be rejected—as Kevin Vallier (2020: 851) puts the case for this: ‘if you include social norms as the subjects of public justification, as Gaus and I have argued, then many antifeminist social norms will be defeated and law may be justified in interfering with defeated social norms to eliminate or suppress them.’ We do not think this is right. Naturally, many would prefer a more egalitarian arrangement. But would they prefer having no rule at all over the inegalitarian rule?

Predictable rules for care work—where we can legitimately demand things from each other—are crucial not only for the creation and upbringing of children and the caring for the old and the sick. It is the fundamental structure upon which all other activity relies. Feminists like Silvia Federici, Mariarosa Dalla Costa and Selma James have powerfully argued that all work outside the home presupposes the care work carried out inside it. Activity must be coupled with passivity; to function in the workplace, workers need rest and relaxation. Our bodies cannot perform the tasks we want them to without being tended to and looked after; our emotional needs must be fulfilled in order for us to be functioning colleagues and workers. Someone must make the bed the worker sleeps in, cook his food and comfort him at the end of the day (Dalla Costa and James, 2017; Federici, 1975). Without the assurance of others’ support, it is difficult, if not impossible, to plan and carry out long term projects and take on commitments. Especially projects whose success involve some uncertainty, but which if they succeeded would be highly beneficial for the community, become increasingly difficult to accept unless there are reliable expectations that the social-reproductive system kicks in should something go wrong (See, for instance, Fraser, 2014). But even in cases where nothing ‘goes wrong’, the social reproduction taking place outside the workplace (in the ‘private sphere’) is essential for the work that takes place in the public sphere (Horgan, 2021).

Whether we fall ill or get pregnant, or simply just have a bad day at work, we want to robustly know that there will be structures in place to care for us. Regardless of how we individuate these issues (that is, whether we should understand them as being governed by the same rule or by one rule each), we need these practices to be rule-governed to work. When evaluating these rules one by one, we believe that it would consequently be commonplace to find that it is much preferable to have each rule over blameless liberty. For instance, a rule that says that women should shoulder a greater responsibility for caring for ageing parents than men, seems preferable to blameless liberty, even for the women who have to shoulder those greater burdens. Because blameless liberty would entail that no one has a moral duty to care for us when we grow old. ¹³ And who would prefer that?

Of course, not having a rule in place for governing how children are to be raised or how the sick and the elderly are to be tended to does not entail that anything goes. Even if people have blameless liberty to do what they want under such an arrangement, they cannot violate other rules in other domains, or basic rights to agency. Gaus's view would not allow that people are left to starve. But it is hard to resist the conclusion that given that they pass some sufficiency threshold, others have blameless liberty to spend the rest of the money they earn on vintage wines, ski trips to Chamonix, and fancy restaurants. To reiterate: under blameless liberty, no one has standing to demand from another that they do (or do not do) something in that particular domain. Women cannot demand from men that they shoulder any responsibilities (except perhaps for providing the absolute minimum), and vice versa. So men also cannot legitimately demand of women to perform the caring roles they would perform under the gendered division of labour.

Clearly, in our societies, women are in many ways made vulnerable and disadvantaged by the rule that says that they should be the primary carer for children, the elderly, and the household. However, absent any rules, they would be even worse off. First, for the reasons just mentioned: it is in everyone's interest that care work is predictably organised. But also because, recall, the rule says that men have a responsibility to provide financially for the family. Under blameless liberty, this positive duty would be lacking. We suspect that in societies with a background structure like ours, many mothers would be in a position where they are likely to ‘pick up the slack’ ¹⁴ absent social norms governing each instance of care work. Thus the standard feminist point about the problem with the ‘double day’ comes sharply into focus: women would do even more of both kinds of labour. ¹⁵

In sum: people would prefer an inegalitarian family norm to blameless liberty in each case. ¹⁶ And if that is true, each of these norms would be in the socially optimal eligible set. Since they are not Pareto dominated, and since they are already established, we have no standing to criticise those who act on them. Thus, the deck is stacked against the disadvantaged. Gaus's view puts them in a double bind (Frye, 1983: 2): if they unilaterally transgress the publicly justified rule, by instead acting on their own preferred alternative, they violate moral obligations (D’Agostino, 2013: 146; Gaus, 2010: 401). If they continue following the rules, their interests are not adequately accounted for.

That said, Gaus asking us to compare a social rule to blameless liberty might be plausible for other domains where blameless liberty could, at least in principle, be acceptable. But it leads to unacceptable implications when the analysis is extended to social reproduction. Since blameless liberty would be so unbearable for most aspects of our intimate relationships, it would still be worse than inegalitarian family norms. ¹⁷

It is important to note that, as we have stressed, many women (and men) would in one sense prefer not to have the inegalitarian rules—they would prefer something else instead. But this is not the test that Gaus uses. The notion of ‘prefer’ at play is rather technical. It is not enough that someone would be better off under an alternative system, and so they prefer that alternative. Or, differently put, the fact that the rule is inegalitarian does not matter for moral evaluation, as long as everyone gains something from it compared to blameless liberty. Gaus believes that there is nothing mysterious about this kind of view. He draws an analogy to the consumer surplus, suggesting that we typically do not think that someone who benefits from it—that is, from paying less than they would be willing to pay for a good—should give up ‘some of his “unfair” gains.’ In much the same way, we should not ask ‘those who are happiest with a moral order… to give up some of their “unfair” gains to those who are less happy with it.’ D’Agostino (2013: 150) believes this to be a reasonable idea on Gaus's part, saying that the fact that

there are any gains at all (over blameless liberty)… depends on the participants arriving at an equilibrium. If they do not, then, not only does not A get his 10 from the coordination on y, but neither does I get her 2. They both get 0 and both have reason to prefer to 0 whatever they would get when they coordinate…

This reveals a deep problem with structuring the theory around the heuristic of asking each Member of the Public whether they prefer a rule to blameless liberty. ¹⁸ Gaus's framework does not allow us to ask of men to give up some of their advantages to women, even if the latter group is structurally disadvantaged by the gendered division of labour, because everyone is at least gaining compared to blameless liberty.

This is objectionable in itself: Gaus's theory cannot support the conclusion that inegalitarian family norms are unjust. But a further implication of all of this is that, by having to compare the status quo to a situation without moral authority, women in societies with background structures like ours are forced to consider what life would be like in a situation without any rule in the domain in question—and thus they have to consider what normally in contemporary moral and political philosophy is understood to be morally arbitrary features as playing a significant role in moral evaluation. This move reinforces women's lower status, based on their hypothetical need for assistance in certain imagined situations. Portraying certain forms of assistance (childcare, midwifery, etc) as something that women should feel grateful for, and constantly reminding women that they would be much worse without the present system, treats women as inferior. It is inconsistent with basic principles of egalitarianism that disadvantaged groups in society, whenever they argued for change, should be reminded that they would be even worse off in another imagined world, and that in this world they would not enjoy any support (support that arguably they are owed anyway). When Gaus's characters Alf and Betty deliberate over whether a rule that Betty finds ‘absolutely unacceptable’ could nevertheless be publicly justified, Alf tries to convince Betty by saying to her that if she rejects the rule, and they cannot agree on another one instead, they ‘will be left with a purely instrumentalist basis for social cooperation… and this will gravely reduce your ability to pursue your ends’ (Gaus, 2010: 312). The implication of Gaus's view, even if it were applied in a society where, as Okin (1989: 171) puts it, one's gender mattered as little as ‘one's eye colour or the length of one's toes’, is that change must go via comparison with a rule-less situation in the domain in question—a state where women are structurally disadvantaged because of morally arbitrary characteristics.

Formal constraints

Defenders of Gaus might suggest that inegalitarian family norms could be ruled out at an earlier stage. If so, the problems we identify would not arise. Gaus lists six conditions—formal constraints—that every rule that is introduced as a candidate basis for a public social morality must meet. The first four are that rules should apply to every person in a given position; be known to all; have some priorities in place for when they conflict; and provide reasons for action. None of these constraints could rule out inegalitarian or exploitative rules.

However, the final two constraints—’Universalisability as reversibility’ and the ‘Modest common good requirement’—might appear suited to do the job. Reversibility states that everyone who proposes a rule must believe that if they were in a different position, they would still accept the rule. The modest common good requirement says that proposals that pose a threat to the good of some Members of the Public are ruled out. This should not be understood in a thick sense—Gaus does not stipulate what the good for all is, because Members of the Public will disagree about such things. However,

we have also come to agree on some basic ingredients of a person's good, such as protection of her bodily integrity, basic liberties to make her decisions about most of the important aspects of her life, control over basic resources needed to live a life, and so on. A rule that severely undermines a person's ability to secure these elements of her good is surely not one that we can reasonably hold she has sufficient reasons to endorse. (Gaus, 2010: 303)

Rules that fail to meet any of the formal constraints are not allowed to form the basis of a publicly justified social morality. So if inegalitarian family norms violate the formal constraints, these rules could never enter the pool of contenders.

We will again grant that rules that are severe violations of bodily integrity are ruled out, also for this reason (however, the line will be difficult to draw—are bans on abortions violations of bodily integrity of the right kind? Are bans on female genital mutilation?). But as we have said, the sorts of practices that we are concerned with need not be harmful in a direct sense. What we disagree with is Gaus's suggestion that even rules that are not violations of bodily integrity could be ruled out at this stage. To illustrate this idea, Gaus introduces the disagreement between John Stuart Mill and his father James Mill over whether women should have the right to vote. James thought that husbands or fathers would see to the interests of women, so there was no need to extend the vote to women. Gaus notes that this judgement probably was sincere—James was not out to oppress women, and he thought that his proposal was reversible (as Gaus (2010: 313) puts it, James probably thought that ‘if he were a woman, he would see the wisdom in his proposal.’) The problem, rather, was that he ‘simply failed to appreciate the good of others.’ As we know, John did not share this view, so when he ‘evaluated his father's proposals in terms of the constraints on moral rules—such as nonoppression—he concluded that they failed’(Gaus, 2010: 314). That is, even though James would have concluded that the rule met the common good and reversibility conditions—and that it was not oppressive—John certainly did not think so. And we ‘cannot accord moral authority’ to rules that fail to meet these conditions (Gaus, 2010: 314).

Naturally, we agree that a rule that deny women the vote cannot have moral authority. But it seems as if Gaus, in order to get to this conclusion, must moralise his account in a way that takes it quite far from the original formulation of his view. How do we know whether James or John is right, without some framework for evaluating their proposals? As Gaus himself stresses, James does not believe his proposal to be oppressive. To make the point explicit, let us suppose that John, instead, proposes to his father that women should have the vote. Would James not simply respond that this rule does not meet the reversibility condition, insisting that if he were a woman, he would not have wanted the vote? And would he not say that this rule failed to properly take into account the good of women? That is, if fathers and husbands are better at taking women's interests into account, why should they not do so? Insofar as this view is sincerely held, which Gaus grants that it is, does that mean that universal suffrage lacks moral authority, simply because James cannot accord moral authority to it? This would clearly be an unwelcome and counterintuitive implication, but we are not sure how Gaus could avoid it without building in some substantive standard of rightness into his definition of the formal constraints. That is, he must make a substantive case for why John is right.

John is right, of course. But since Gaus emphasises that moral evaluation always must be relative to each Member of the Public's own evaluative standards, how can we, if we limit ourselves to his framework, ‘know’ this? If Gaus moralises his account such that only those rejections based on formal constraints that are ‘correct’ are permissible, this account starts to look like a very different kind of moral theory. Because then it seems as if Jane could suggest that rules that, say, only allow two weeks of parental leave, privatises health-care expenses, or ban abortions, also violate the nonoppression or common-good conditions. If so, the only rules that we would be allowed to choose from would be the rules that are in line with Jane's view of egalitarian justice. Some people would dispute this, of course, but so does James vis-à-vis John's proposal. Hence, that fact alone cannot block this move.

We take it that Gaus would not want to open this door. But if John is allowed to appeal to formal constraints, and James is not, then the onus is on Gausians to explain why Jane is not allowed to do the same thing. Is it because she is wrong about what the formal constraints entail? That would require an explanation of the asymmetry between her views and John's—why is John right and Jane wrong? It seems to us that if Jane is not allowed to reject proposals based on the formal constraints in these cases, then neither should John be allowed to, unless we have some way of saying that John is right and Jane is wrong. Absent such an argument, it appears as if Gaus must choose between saying either that no rule would be publicly justified if there is disagreement over the formal constraints, or reverting to the main feature of the theory, by saying that if having a rule is better than the baseline—’blameless liberty’—then the rule is within the set of permissible rules. And clearly it must be better to have suffrage only for men than having no ‘peaceful’ mechanism in place for resolving political disputes. Hence, if these are the right implications of Gaus's view, either no rule could ever be justified, so we get to anarchy after all, or we would end up with a publicly justified rule saying that women should be denied the vote. Bitter pills to swallow either way.

Patriarchal origins

Some readers might wonder whether the fact that the inegalitarian rules so obviously have a patriarchal genealogy might make a moral difference. There is indeed a common intuition that once our beliefs and practices are revealed to have a certain origin, this casts doubt on whether we are justified in holding on to them. The patriarchal roots of the rules we have discussed could in themselves induce a form of genealogical anxiety, that might lead us to question their legitimacy. ¹⁹ Moreover, one could suspect that the preferences of those who live under patriarchal systems might, plausibly, change and distort over time to align with patriarchal norms. Should we really just take these adaptive preferences as they are? Taken together, the reader presses, does this not imply that Gaus could escape our charge?

In response to these queries, it is important to note that Gaus does not consider a rule's origin to be morally relevant. He sees, as it were, no grounds for our genealogical anxiety. He grants that many rules have genealogical problems, but this does not change their moral status. As Gaus (2010: 416) puts it, ‘many of our current rules were simply imposed on society by the powerful; one would have to be extraordinarily naive not to appreciate that features of our positive morality have been shaped by unjustified power.’ Nevertheless, this is not relevant for moral evaluation, because ‘what is ultimately important are not efficient causes but sustaining reasons’ (Gaus, 2010: 416). Hence, even if the rules are imposed by the powerful, those who do not belong to that group might still come to endorse them.

Similarly, the fact that preferences are ‘adaptive’ does not make them invalid for Gaus. ²⁰ Indeed, he makes the opposite point. In the real world, people—as a matter of empirical fact—come to accept the social arrangements that they live under. Those who were ‘initially critical’ of a ‘system of rules’ might ‘come to embrace it’. Adaptive preferences, for Gaus (2010: 408–9; see also D’Agostino, 2013: 150), thus ‘is a great and important thing, not a cause of regret,’ because it is ‘a crucial source of moral stability.’ Again, what one might have thought of as a bug turns out to be a feature of his theory.

We disagree with Gaus on this. As D’Agostino (2013: 150–151) notes, the fact that we can ‘become adapted to a situation in which, as an inevitable consequence of some individuals having their values fulfilled to a very high degree by our arrangements, we have ours fulfilled to a significantly lesser degree, is certainly handy and, indeed, “a crucial source of moral stability,” but it is also very convenient for the “privileged” in this situation.’ It has been known at the very least since the days of Aesop's foxes and sour grapes that our preferences are easily adapted to how likely it is that we can meet them. The fox may have reason to value his ability to adapt, but where there are background injustices, and where some groups are assigned a lesser share of certain goods and privileges, this tends to affect in an objectionable way the beliefs and preferences of their members. ²¹ As Ann Cudd (2006: 181; see also Khader, 2011) puts it,

if one's preferences adapt to oppressive circumstances, then one's desires turn away from goods and even needs that, absent those conditions, they would want. Oppressed persons come to see their conditions of oppression as the limits within which they want to live.

So, where women are encouraged to wear high heels, they will tend to adapt their preferences and beliefs so that they will find it good and attractive to do so (Walsh, 2015: 834). Or in social systems where women are told (by social norms, the state, religious sources, and so on) that they are only or mostly fit for doing work outside the public sphere, women will tend to believe this, and come to prefer such work. Cudd (2006: 181), again: ‘It is not that they will prefer oppression to justice, or subordination to equality, rather they will prefer the kinds of social roles that tend to subordinate them, make them less able to choose, or give them fewer choices to make.’ And as Amia Srinivasan (2020: 399) puts it, what someone in this situation believes

is not merely false, but moreover the product of a convincing, and systematic, patriarchal illusion: that it is men's place to subordinate women. This illusion—one that infects not only the testimony of [her] peers and respected elders, but her moral emotions (shame, remorse) and best attempts at rational reflection— ensures that [she] has no dependable access to the moral facts of her situation.

The problem is not that Gaus fails to acknowledge the patriarchal origins of many of our social rules and the resulting adaptive preferences, but that he draws the wrong conclusion from these facts. While feminists emphasise that adaptive preference formation as a consequence of social expectations gives us reason to be suspicious of accounts that rely heavily on the fulfilment of desires in their evaluation of social systems, Gaus celebrates it as a phenomenon that allows social systems to remain stable over time. ²²

The best we can hope for?

There are a few ways in which Gaus might respond to our arguments. For instance, he could maintain that it would be invasive for the state to do anything about inegalitarian family practices. The thought would be that in insisting on the liberal divide between public and private, we protect people's autonomy and their privacy. In that sense, it might be regrettable that the theory delivers the judgement that inegalitarian family structures are permissible, but it would be objectionable to remedy this with state power, because doing so would constitute an even greater injustice.

In response, it is important to note that our view, as presented here, does not commit us to any one way of mitigating these injustices. Whether or not the state should intervene is beside the point. ²³ We are not discussing law or even public policy. Instead, what is at stake is our standing to blame those who engage in patriarchal family practices. They are acting in a morally permissible—if not obligatory!—way. Indeed, Gaus's theory seems to imply an even stronger claim. The person who rejects a publicly justified inegalitarian family practice by demanding from her partner that they split care work equally is the one who is committing a moral wrong.

Gaus could say that people have legitimate expectations to continue organising their lives after the social norms and practices that they have so far already accepted. Gaus (2010: 425) might say that his theory ‘privileges the status quo,’ and for good reason. On Gaus (2010: 426) view, there is a ‘relevant difference to which our moral theory must be sensitive’ between the status quo and competing alternatives. Hence, if we

compare two rules in the optimal eligible set, one that is now currently acknowledged and one that is not, only the former does what we require a moral rule to do: provide an authoritative basis for interpersonal moral demands (Gaus, 2010: 427).

We are, then, not ‘simply “preferring” the status quo’, we are ‘choosing an option that has the crucial, additional, moral property of being an equilibrium such that all have reason to act on it, and so, uniquely, I treat you as a free and equal moral person when I demand that you conform to it’ (Gaus, 2010: 426). Moreover, we cannot be certain that rejecting the status quo will in fact result in something better.

We agree that people do have legitimate expectations in relation to social reproduction and family life. There would most likely be severe consequences if society were to overturn its social norms in this area overnight. However, Gaus has not convinced us of how or in what way this would happen—and even if we were so convinced, we have not suggested we would want that. We are not suggesting anything about how reform should be sequenced, and it might be that we have prudential reason to prefer piecemeal reform. ²⁴

Conclusion

At different points, the option for Gaus to simply bite our various bullets clearly is available. Let us therefore end by providing a final reason for why these bullets would be particularly unpalatable for those who, like Gaus, are concerned with public justification.

Any plausible moral theory must be able to find a way of balancing the expectations we have of others with the possibility of moral change and reform. We have argued that Gaus's theory fails in the second regard by putting an undue emphasis on the former. But from a public justification perspective, there is also a further question about whether we could really say that a practice is publicly justified, and justified to me.

We think that this test does not seem to be met, if after careful consideration I find that I am significantly burdened by the rule, and that there are available alternatives that would be much better for me. We therefore believe that, even from within Gaus's own standards of ensuring freedom, equality, and public justification, the framework provides an unsatisfactory test. That is, the ‘blameless liberty’ baseline is a mistake, and we need some other test for whether a rule is justified to someone. Inegalitarian rules cannot truly be justified to each and every one of us.