2. The normative background

As in the case of other scarce goods, we need methods to distribute health-care resources. Such distributions can be better or worse, morally speaking;Footnote ³ a widespread conception about how we ought to distribute health care is prioritarian, attaching more normative weight to benefits that accrue to the worse off recipients. As distributive devices, ordeals have been primarily praised as good tools for reducing wastage and directing resources to the most needy recipients (Altas et al. Reference Altas, Banerjee, Rema, Benjamin, Purnamasari and Wai-Poi2013; Olken Reference Olken2016; Sunstein Reference SunsteinForthcoming); if the chief virtue of well-designed ordeals is that they result in more efficient, need-tracking distributions, this makes them desirable in terms of either utilitarian or prioritarian goals. I favour the view that prioritarianism is the right principle for distributing goods the distribution of which is a matter of justice. Alongside health care, goods whose inclusion in the distribuendum of justice is relevant to the arguments of this paper are not only the generally acknowledged income, wealth and opportunities to positions of advantage (Rawls Reference Rawls1993), but also discretionary time (Goodin et al. Reference Goodin, Rice, Parpo and Eriksson2008; Rose Reference Rose2016).

Some ordeals are likely to have undesirable distributive consequences both within and outside the area of health care. In particular, I shall explain the possible disproportionate negative impact on some groups of women in terms of their access to health-care resources, on their discretionary time and, in some cases, on their opportunities in the labour market. On the other hand, distributing health-care resources by ordeals rather than according to ability and willingness to pay are likely to benefit the (monetarily) poorest women. Women on average are worse off than men both in money and in discretionary time (Goodin et al. Reference Goodin, Rice, Parpo and Eriksson2008). Hence, ordeals can improve on gender justice, at least relative to cost-sharing.

Before I expand on each of these claims in the subsequent sections, let me make some stipulations about what counts as setting back, or advancing, the goals of gender justice, and who the agents are that may, or even have a duty to, advance the goals of gender justice.

As I understand it, and already mentioned above, gender justice encompasses two aims. The first is to undo gender itself – to eliminate gendered norms from children’s socialization and from the workplace, and to dispel the gendered assumptions that shape economic life; a prominent example of such assumptions is that ideal workers have a partner who relieves them from domestic and care responsibilities. The second aim is to rectify those inequalities of outcome between women and men which are due to formal or informal gender norms. I explain these two aims in turn.

One feature that makes gender norms objectionable is their arbitrary influence on people’s development of talents and ambitions (Okin Reference Okin1989; Gheaus Reference Gheaus2012); some feminists go as far as criticizing gender norms for impairing women’s autonomy (Chambers Reference Chambers2004). In addition to stunting individual development, gender norms are responsible for unfair disadvantages. Even in economically advanced liberal democracies, women reach worse outcomes than men with respect to lifetime power, status, attachment to the labour market, equal pay for equal work, the holding of better jobs and discretionary time. The worse-affected groups of women, at least in terms of economic power, are those who have children (Schouten Reference Schouten2019). Single parents and women who, over the course of their lives, have participated little, or not at all, in the labour market are at particularly high risk of poverty and marginalization, even if they dedicated their time to caring for family and friends (United Nations 2010). Moreover, factors such as statistical discrimination and explicit and implicit sexist biases together undermine equality of opportunity between women and men for social positions of advantage (Gheaus Reference Gheaus2012).

Feminists frequently argue that the main culprit for these inequalities is the gendered division of labour (Okin Reference Okin1989; Gheaus Reference Gheaus2012; Schouten Reference Schouten2019). Women’s tendency to specialize in providing care for family members, and men’s to specialize in paid employment fuel statistical discrimination against women qua workersFootnote ⁴ as well as less rational forms of bias. Moreover, since markets reward an ideal worker who is supposed to be free from care duties (Fraser Reference Fraser1994; Williams Reference Williams2000), women who specialize in care-giving at home have few opportunities to also hold decent jobs or to resume attractive careers later in life. Thus, they often become economically dependent on their wage-earning partners. Many families rely on all adults participating – at least to some extent – in the labour market; in such families women’s and men’s tendency to specialize in either care-giving or paid work is also responsible for the fact that, on average, mothers enjoy significantly less discretionary time than fathers (Rose Reference Rose2016). This is because women tend to have part-time jobs as well as being responsible for domestic chores and care-giving.

Such women also enjoy, of course, less discretionary time and economic opportunities than women and men who are not care-givers. Given the extent to which the care-giving responsible for women’s lesser advantage is directed to children, the gap in advantages enjoyed by care-givers and by non-care-givers is not an obvious injustice. Whether or not it is, depends on the question of whether the costs of rearing children ought to be supported by parents alone, or whether all members of society ought to share them.Footnote ⁵ This is a debate with which I cannot engage here. Assume that justice requires that the costs of rearing be shouldered by parents alone, at least in cases in which parenthood is voluntarily undertaken. Even in that case, the gap in discretionary time and economic opportunities between mothers and fathers is unjustifiable at the bar of justice, since both mothers and fathers have the same parental duties and it is unfair if fathers free-ride on mothers voluntarily discharging most of these duties. And even if parents alone ought to shoulder the costs of rearing, some women who are not mothers will be disadvantaged by the gendered division of labour: implicit biases against women as workers, or stereotyping women as more fit for domestic roles than men operate against women in general. The gendered division of labour disadvantages women in general, although to different degrees depending on whether they are also parents.

Inequalities in outcomes between women and men are condemned by gender justice to the extent to which they are the result of gender norms. Often, such norms are operational as part of women’s (and men’s) socialization or as generating unfair constraints on women’s participation in the market, as is the case with employers’ and co-workers explicit or implicit stereotypes against women. But there are also cases when past gender norms generated unrectified historical injustices responsible for path-dependent institutional developments. Examples of the latter include former legal obstacles to women’s education, employment and holding of property, which helped create an economic model that assumes care-givers either stay at home or work part-time;Footnote ⁶ according to this model, ideal employees have a level of commitment to their jobs that is incompatible with them being part-time care-givers. This model interacts with current gender norms, and the result is that the vast majority of people who sacrifice market participation in order to provide care are women.

Now, it is of course impossible to identify with any precision which differences of outcomes are due to past or present gender norms, and which to behaviour that would also exist in the absence of any gender norm. But, given the pervasive nature of gender norms, some philosophers conclude that it is safe to assume that all significant inequalities of power, status and wealth between women and men are objectionable at the bar of justice (Phillips Reference Phillips2004).

The first aim of gender justice is to eliminate unjustified gender norms. One of the reasons behind this aim is to give women and men the same opportunities for advantageous outcomes, which would involve the elimination of gender norms that curtail women’s economic, political and social aspirations, and that encourage heterosexual men, but not heterosexual women, to rely on their partners for the provision of care.Footnote ⁷ It would also involve the elimination of gender norms that prevent women’s fair treatment in the workplace. Indeed, some feminists think that the elimination of obstacles to full equality of opportunity between women and men is the only proper aim of gender justice and that the concern with unequal outcomes is confused (Radcliffe-Richards Reference Radcliffe-Richards2014). But gender norms are pervasive and not easy to eradicate. As long as they are operational, the inequalities in outcomes they generate are also objectionable. The second aim of gender justice is to prevent or mitigate inequalities of outcome generated by gender norms, such as the ones explained above.

It is not always possible to pursue these two different aims of gender justice via the same policies; moreover, some of the policies that seek to mitigate the inequalities of outcome generated by gender norms inadvertently entrench gender norms. This difficulty is reflected in the feminist debate about whether gender justice requires societies to move towards a more gender-neutral division of labour or whether it requires that societies accommodate and support those women who perform the large bulk of care-giving (Robeyns Reference Robeyns and Browne2007; Gheaus Reference Gheaus2012; Schouten Reference Schouten2019). The latter strategy enables women to continue providing most care-giving without facing economic and other penalties. It may be impossible to satisfy both goals at the same time, since supporting care-givers is likely to provide incentives for a sizeable group of women to continue to specialize along traditional gender lines; indeed, sociological research indicates that a significant number of women report a stable preference for combining care-giving with part-time work (Hakim Reference Hakim2000). As long as a critical mass of women specialize in care-giving (and no similar trend exists amongst men), statistical discrimination against women and other biases are likely to endure. For the purpose of this paper I assume that, in cases of practical tensions between these two aims of gender justice, priority should be given to the interests of the worst-off women.

When policies entrench the gendered division of labour without compensating women, for instance by exploiting the ways in which women are made vulnerable by their specialization in care-giving, or by worsening women’s time-poverty relative to men’s, these features make them pro tanto objectionable at the bar of gender justice. Some ordeals may be objectionable for precisely these reasons. But the objectionability of a policy that entrenches gender norms may be mitigated if at the same time it compensates for some inequalities of outcome generated by gender norms, especially if these inequalities affect the worse-off women. As I explain below, time-consuming ordeals are likely to entrench gender norms but also to advance the interests of the worse-off women relative to cost-sharing alternatives.

Further, there is an old and ongoing debate amongst feminists about the agents that can legitimately combat gender norms and the gendered division of labour. Some philosophers who take the gendered division of labour to be the main cause of gender injustice have argued that the family is part of the basic structure of society, and hence comes under the purview of justice and of legitimate state intervention (Okin Reference Okin1994). Yet, many people endorse traditional, gendered, lifestyles as part of their reasonable conception of the good. Political liberals, who believe that states ought to remain neutral with respect to their citizens’ conceptions of the good, take this as a reason against state-mandated, gender-egalitarian interventions (Rawls Reference Rawls1993; Lloyd Reference Lloyd1995). More recently, feminists defended gender-egalitarian policies against the charge that they are incompatible with political neutrality by pointing to various desiderata of justice: autonomy (Chambers Reference Chambers2004; Schouten Reference Schouten2019); equal capabilities (Robeyns Reference Robeyns and Browne2007) or opportunities (Gheaus Reference Gheaus2012) between women and men; equal citizenship (Hartley and Watson Reference Hartley and Watson2018). Assuming that at least one of these defences is successful, then it is legitimate to use ordeals – as well as other policies – as incentives for individuals to adopt more gender-egalitarian divisions of labour.

3. Effects of ordeals on women’s time and access to health care

The most obvious effect that some ordeals might have on gender justice is related to the relative command that women and men have of resources such as mental health, time, education, or self-confidence.

A first concern here is that women tend to display, to a larger extent than men, some features that can prevent them from accessing health-care resources distributed by ordeals. For example, depressed people are likely to find it particularly burdensome to wait in long lines (Eyal et al. Reference Eyal, Romain and Robertson2018: 16) or to do additional paperwork, and prone to be disproportionately, and negatively, affected by ordeals that involve queuing or form-filling. Since women are affected by depression almost twice as much as men (Albert Reference Albert2015), ordeals of this shape will set back women’s interest more than men’s. Serious depression – let alone depression in general – is a condition that affects a significant number of individuals, and hence the effect is noteworthy.Footnote ⁸ Note that the setting back of interests can be in terms of deterring women, to a larger extent than men, to access the health-care resources distributed by ordeals of this kind; or, else, it can take the form of making it psychologically more costly for women than for men to access the health-care resources. This is obviously an issue of gender justice to the extent to which women’s depression is caused by gender norms that result in women’s lesser power and status, or pressures to combine paid employment, housework and care or, indeed, norms that condone violence against them. But even in the absence of such relation, the overburdening of women by ordeals is problematic. Assuming that women, on average, are worse off than men, and assuming that women who suffer from depression are a particularly vulnerable group, ordeals that prevent, or make it disproportionately difficult for, women to access resources, worsen women’s situation and hence raise a worry of gender justice.

A similar analysis applies to the time gap between women who are primary care-givers (as well as employees) and other social groups. Time-consuming ordeals have been defended as a good means to offset the disadvantage that poor people have in accessing health resources, because ‘whereas money is particularly scarce for the poor, time is not, and so time costs can be used to screen’ (Olken Reference Olken2016: 865). But this is not equally true of women and of men. For women, discretionary time is a scarcer resource than for men whenever they work more hours per week than men – that is, when they do a ‘second shift’, for instance by combining participation in the labour market and shouldering the lion’s share of domestic burdens (Hochschild and Machung Reference Hochschild and Machung1989). For women, who have less discretionary time, minutes have higher non-monetary marginal value.Footnote ⁹ Distributing health care by ordeals is typically more time-consuming than using other methods, and therefore ordeals can introduce a relative disadvantage for women with very little discretionary time compared with men (or, indeed, compared with women who are relatively well-off in terms of discretionary time). Moreover, the gendered division of labour means that women are likely to deal with ordeals not only on behalf of themselves, but, when the ordeal design permits it, also on behalf of other family members. This is an additional way in which ordeals in health-care distribution are likely to put a disproportionately large burden on women who are care-givers. Women in general have less discretionary time than men and mothers in mother-led households in particular are the worst-off in terms of discretionary time poverty (Goodin Reference Goodin2011). So, the same argument that recommends ordeals on grounds of class – because ‘money is relatively scarcer than time for the poor’ (Olken Reference Olken2016: 865) – may dis-recommend them on grounds of sex.

Inequalities in discretionary time are, indeed, likely to generate important complaints of justice, especially if the worse off in this respect fall below a minimum threshold of adequacy with respect to discretionary time. Recently, several political philosophers have argued that having enough discretionary time is a central component of freedom, and/or one of the goods the distribution of which is a direct concern of justice (Goodin et al. Reference Goodin, Rice, Parpo and Eriksson2008; Rose Reference Rose2016). Further, one of the primary concerns of gender justice is with the way in which gender norms, and the gendered distribution of labour, exclude women from full citizenship (Hartley and Watson Reference Hartley and Watson2018); obviously, one of the necessary conditions for participating in social and political life is to have enough discretionary time. More generally, sufficient discretionary time is needed for individuals to get the fair value of their basic liberties (Rose Reference Rose2016): having rights to movement, free association or speech cannot mean much to individuals who lack the time to exercise these rights.

Finally, inasmuch as rationing by ordeals happens in parts of the world where women lag behind men in terms of literacyFootnote ¹⁰ – that is, usually not in advanced liberal democracies – those who design them ought to also pay attention to this education gap. Ordeals that require literacy, and confidence in one’s literacy, can, in some circumstances, exclude or burden women significantly more than men.

Women for whom ordeals would be particularly burdening due to mental health issues such as depression, or to insufficient education, are also likely to be (in virtue of the same features) amongst the worst-off members of their societies. The same applies to women who are very poor in discretionary time because they are single parents and/or because they need to hold several jobs to make ends meet. (By contrast, women who are time poor because they have demanding careers as well as, possibly, caring responsibilities, are not amongst the worst off, although, compared with their male peers, they are on average unfairly disadvantaged.) This means that negative consequences of ordeals on these groups of women have much normative weight; if, on the balance of reasons, time-consuming ordeals are found desirable, policymakers ought to try and offset the burdens that ordeals are likely to place on women who suffer from depression, illiteracy and time-poverty due to economic hardship. If discretionary time, mental health, and health care are part of the distribuenda of justice, worsening women’s access to these goods generates a complaint of justice.

In spite of such considerations, in some cases ordeals may be the most desirable mechanism in the allocation of health-care resources (either from the point of view of gender justice, as I explain in the next section, or even all-things-considered). When ordeals are adopted in such circumstances, considerations of gender justice indicate that they should be coupled with ameliorative measures to minimize their negative impact on women’s time and (access to) health.Footnote ¹¹ I illustrate such a possibility in the next section.

4. Ordeals and opportunities in the labour market

Another type of plausible effects of ordeals on gender justice is less direct, because it doesn’t necessarily impact on the differences in aggregate outcomes between women and men. Instead, this potential effect is particularly relevant for the first aim of gender justice explained above because it concerns the way in which rationing by ordeals can entrench gender norms. On the assumption that women are generally expected to deal with time-consuming ordeals on behalf of some of their family members, the resulting loss of their time will have a negative impact on their opportunities in the labour market, and, more generally, on their opportunities to engage with public matters. This will set back their opportunities to positions of advantage (and hence worsen inequality of opportunities between women and men); but the same effect will also buttress the gendered division or labour, keeping women away from economic and political power.

Here is the explanation of this mechanism. People who take responsibility for meeting the needs of others – their children, sick, elderly or disabled family members, and friends – are, other things equal, less available to their employers, need more time flexibility in their job, and are more likely to use their leave entitlements; thereby, they fall short of the ideal worker. If ordeals contribute to this disparity between women’s and men’s behaviour on the labour market they also contribute to the reinforcement of statistical discrimination, because they strengthen employers’ reasons to expect female employees to require more leave and flexibility than male employees. The effects of ordeals alone is probably minor when compared with the bulk of women’s systematic, long-term care responsibilities for children, ill or disabled relatives, or to their general household management responsibilities. Yet, it seems important to register this kind of complaint of gender justice against time-consuming rationing methods.

However, the fact that some ordeals can contribute to setting back equality of opportunity between women and men need not mean that introducing them has, all things considered, undesirable consequences for gender justice. Unlike the women afflicted by depression, or who lack adequate education to manage ordeals, or who are time-poor for reasons of economic hardship, women whose market opportunities are most affected by statistical discrimination are not amongst the worst-off members of their societies. Statistical discrimination is more likely to operate against women who aspire to attractive jobs: jobs that are well paid and display other attractive features, such as work autonomy, security, interesting tasks and good future prospects, usually require high qualifications; they also tend to demand workers’ commitment to full-time careers. Other things being equal, economic rationality speaks against hiring for these positions workers who need flexible working times. It is also for these kinds of jobs that employers want to avoid high levels of staff turnover – for instance, due to parental leave – since finding adequate replacements and training new employees is costly. In contrast, the jobs where a high level of staff turnover is not particularly problematic for employers – because it is easy to recruit replacements, and cheap to train new workers – come with significantly lesser benefits. Therefore, ordeals that worsen women’s time-poverty will mostly fuel statistical discrimination against women who aspire to the better jobs – that is, women who are amongst the best-educated and skilled members of their societies, and hence not amongst the worse-off.

Assume that the alternative to rationing health-care resources by ordeals is to ration them by cost-sharing, and that in some cases rationing by ordeals is, on the whole, better for poor women than rationing by cost-sharing. This can be the case whenever women’s interest in more disposable time (set back by ordeals) is outweighed by women’s interest in more cash (set back by cost-sharing.) In such cases, policymakers face a trade-off: they can opt for ordeals and impose burdens on the better-off women by feeding the reasons for statistical discrimination against them, and thus also entrenching gender norms. Or they can opt for cost-sharing and impose disproportionate financial burdens on the worse-off women. Ordeals are one case of policy in which advancing one goal of gender justice – the undermining of gender norms – is likely to set back the other – compensating women for their (in this case, material) disadvantage.Footnote ¹²

To illustrate, take a stylized example: Jane and Joe are a working class couple, Mady and Max are a middle class one. In their respective households, it is Jane and Mady who deal with ordeals – they would wait in lines for themselves and maybe family members, fill in the extra forms and so on. This means that Jane and Mady would pay the costs of ordeals in terms of time and effort. Rationing through inconvenience would set back Mady’s interest in time. It would increase the demand for care leave and flexibility time from her, and those like her. As a result, her opportunities relative to men’s for attractive jobs would be further set back – perhaps just a tiny step – by rationing via ordeals. Since Mady aims to have a career, and has a real shot at an attractive job, time is a high price for her to pay because it translates in loss of high-stakes opportunities. Ordeals would also burden Jane – who is also more time-poor than John – by reducing her discretionary time; but they would be less likely to affect her opportunities on the market, since she is more likely to have a precarious job, not a career. Jane will probably also need leave from her job, but, unlike Mady, she will not miss important meetings on which her career depends, and if the leave is longer her employer will find it easier to replace her than Mady’s employer will. Jane’s employment conditions are, in any case, likely to be more precarious than Mady’s. Since Jane has little money, she may be actually better off with less time (the price of ordeals) than with the money she would have to pay if necessary health-care resources were distributed by cost-sharing. By contrast, cost-sharing would require no time sacrifices from Mady, and hence it would not negatively impact her career. Cost-sharing would thus be better for Mady’s opportunities in the labour market, but it would obviously set back Jane’s (and Joe’s) financial interests more than Mady’s (and Max’s), who are better off economically and therefore are less burdened by cost-sharing.

One way out of this dilemmaFootnote ¹³ would be to distribute health-care resources by cost-sharing that is adequately progressive, such that the financial burden imposed on people like Jane (and Joe) is less significant than the time cost imposed on them by an ordeal on both women but mostly on Mady. But if, for whatever reason, the dilemma cannot be dissolved,Footnote ¹⁴ and if policymakers opt for ordeals in order to prioritize the interests of the worse-off women (and men), it is important to be aware of the higher costs it imposes on women and on the different costs to be paid by the different groups of women.

Women such as Jane are likely to be negatively affected by time-consuming ordeals because their discretionary time will be further diminished. If they opt for ordeals, policymakers can mitigate this effect either by providing incentives for men like Joe to share the time-burden of ordeals – and thus, by undermining gender norms – or, more feasibly in the medium term, by making it possible for women like Jane to spend less time in paid work. Such solutions, like housewife wages, have long been advocated by feminists (Okin Reference Okin1989) as a general way of compensating women’s care work.

Ordeals will affect women like Mady not only by making them more time poor but also in terms of their opportunities in the job market, and they will aggravate the inequalities of such opportunities between women and men. Even if this cost cannot be eliminated through the design of particular ordeals, it may be possible to mitigate it by introducing other policies that combat statistical discrimination and, more generally, biases against women in the workplace.Footnote ¹⁵ Even more importantly, policymakers can provide incentives to people like Max to share informal care-giving more equally (Schouten Reference Schouten2019).

5. Designing ordeals to advance gender justice?

The use of ordeals in rationing healthcare could perhaps be turned into an asset for gender justice. For instance, ordeals can be designed to include defaults meant to improve a more equal participation of women and men in child-rearing by nudging men to get more closely involved in the provision of healthcare for their children – starting before birth.

As already noted, many feminists think that gender justice requires moving towards a more gendered-neutral division of labour. De-gendering the division of labour may be instrumentally necessary to achieve gender justice, as the best or the only way to undermine gender norms, including those that prevent equality of opportunities between women and men (Robeyns Reference Robeyns and Browne2007; Schouten Reference Schouten2019). On an alternative view, a gender-neutral division of labour is the regulative ideal of gender justice as a necessary (although not also sufficient) condition for giving both women and men opportunities to combine care-giving with secure participation in the labour market (Fraser Reference Fraser1994; Gheaus Reference Gheaus2012). In both cases, in order to de-gender the division of labour it is important to get men more involved in care-giving.Footnote ¹⁶ One efficient way to involve men in care-giving is to encourage them to step into their parental role from the beginning of the child’s life (Brighouse and Olin-Wright Reference Brighouse and Olin-Wright2008) – or perhaps even before birth. Indeed, for this reason feminists proposed ways of reshaping, for instance, parental leave, to incentivize fathers to spend as much time with their newborns or adopted children as mothers (Brighouse and Olin-Wright Reference Brighouse and Olin-Wright2008; Gheaus and Robeyns Reference Gheaus and Robeyns2011). If, as a default, both expecting parents were required to participate in the delivery of health care for pregnant women during pregnancy, this might result in better male participation in parenting. Requiring, as a default, custodial fathers to participate in some of the delivery of health care to their children may advance the same goal. At a minimum, health-care providers could require expecting fathers who do not wish to participate in the delivery of pregnancy care to fill out forms explaining the reasons for their unwillingness to participate. Usually, acknowledgement of paternity happens at birth; however, some countries allow men to acknowledge paternity before birth, for instance, when the expecting couple is not married, and in order to make sure paternity was acknowledged in case the expectant father dies before the birth of the child. This practice could be extended to all heterosexual couples when the pregnancy is established, and be the occasion of additional form-filling for fathers who decline to be part of the pregnancy health care. A similar default could apply in order to encourage fathers to take charge of some of the delivery of health care to their children. Policymakers could go a step further and use nudges such as shorter waiting times if fathers, rather than mothers, take their children to the doctor.

The design of any ordeals that include nudges towards a more equal involvement of fathers in child-rearing would need to be especially careful about any unintended consequences. In particular, policymakers ought to make sure that the design of defaults does not set back children’s interests (by alienating instead of involving fathers), does not backfire on expecting mothers who don’t have partners (by stigmatizing them) and does not overburden expecting mothers who have partners unwilling or unable to participate. An artfully designed ordeal would encourage men’s participation in the care of their pregnant partners and children without setting back either women’s or children’s access to medical resources.

In particular, it has been argued that using defaults which assume that both parents are going to perform the same amount of child-care is a good compromise between nudging individuals towards more gender-just arrangements and preserving their freedom to decide (with some inconvenience) to follow less symmetrical patterns of care-giving. Policymakers can do a similar thing in the case of ordeals rationing pregnancy care and care for children.