‘Flexible’ Family Leave Is Lousy Feminism

“You must know lots about Denmark,” she says. 

I’ve just told my new acquaintance that I’m a political philosopher, writing about family support policies.

From that unmistakable longing in her eyes I’m guessing she’s a mother. Maybe she longs for paid leave or help managing the overwhelming cost of daycare. Maybe she wants to enlist her partner as a genuine co-parent, rather delegating discrete tasks to him. (Maybe she longs to stop longing for that, because it only leaves her frustrated and because she’d prefer to feel simple gratitude for all that he does do.) Maybe this woman has no parenting partner. Maybe she just wants a better way to get through the week. Or a workplace where she doesn’t have to pretend that dinnertime meetings aren’t a problem for her. Or an employer who doesn’t require workers to drain their sick and vacation time when they take leave under the Family and Medical Leave Act.

I later learn that sleep tops the list of things this mother is longing for. She thinks Denmark’s generous paid family leave and childcare subsidies could make her load a little lighter. I know that longing—I’ve felt that longing, though I have little right to it—and I don’t want to disappoint her. So when she asks whether I think we should replicate Denmark’s family support regime in the United States, I have to steel myself to confess: 

“Well, not exactly. . . . ”

It’s not that I don’t see the problem Denmark’s policies are supposed to solve. Parents, especially mothers, are doing too much with too little support—emphasis on the “especially mothers” part. Women still do far more childcare than men. Meanwhile, labor markets in the United States reward those who can commit to working long hours with little time off, penalizing those with serious childcare responsibilities. A growing body of research confirms that the gendered pay gap persists mostly because women take on a larger share of responsibility for childcare and because they make employment decisions to accommodate it.

We should not be optimistic that large-scale change will come by way of the kind of intra-industry competition that has improved family-friendly offerings in certain high-status professions. While highly skilled and highly sought-after employees can bargain for family friendly policies, employers won’t bend over backward to accommodate workers they consider more expendable or replaceable. Competition among employers won’t help parents outside the workforce, either. It won’t fix our collective failure to recognize caregiving work as work, nor (relatedly) our tendency to under-value it. Meanwhile, it will drive demand for professional caregiving services without generating any coordination mechanism to ensure that caregivers are compensated fairly. And competition among employers won’t help involuntarily unemployed men contend with norms of masculinity that tie their social worth to their status as earners. Gendered norms about caregiving and work are the linchpin for a lot of social ills, and labor markets on their own won’t solve the problem. We need social change that only politics can deliver. We need social policy.

So why oppose Danish-style paid leaves and subsidized childcare? It’s simple but perhaps counterintuitive: Danish-style policies support caregiving but won’t move the needle on gender inequality.

Don’t get me wrong. A lot of the policies that get most attention in this arena should be part of any social policy solution. Work-time regulation—especially to spur increased flexibility and decreased time demands—is essential. And childcare support of some kind must be part of the solution. The problem with Danish-style childcare support is that it’s too flexible. It helps women balance work-life pressures, which should be one important aim of policy like this. But men can still opt out. And they do. To move the needle on gender inequality, family support policy must be crafted to influence who uses its provisions and how.

In other words, we need family support policy that is deliberately inflexible and intrusive: policy that is explicitly designed to incentivize childcare among fathers and labor market attachment among mothers. The main hurdle to justifying such policy is a philosophical question of legitimacy: May the state legitimately undertake such coercive policy? 

I’ll answer yes. But first, why doesn’t flexible policy work?

---

Like a great deal of other research, a 2018 study using Danish administrative data finds that earnings inequality now mostly reflects different choices women and men make about how to care for children: women, more than men, “favor family amenities over pecuniary rewards,” and suffer an earnings penalty as a result. Throughout almost all of the thirty-year period studied, Denmark provided generous job-protected paid parental leave so that parents could care for their children at home during their first six to twelve months. After the paid leave period, parents gained access to high-quality and heavily-subsidized options for purchasing care outside the home. Even so, significant gender inequality persists:

while gender inequality in Denmark was dramatically lower than in the United States around 1980, today the gender pay gap is between 15–20 percent in both countries and appears to have plateaued at that level. That is, gender convergence happened earlier in Scandinavia than elsewhere, but the process also slowed down earlier in Scandinavia allowing other countries to catch up. So even though these countries feature different public policies and labor markets, they are no longer very different in terms of overall gender inequality.

Although Denmark’s generous family support provisions nudged them toward pay parity, the United States—which offers no paid leave, and where the costs of daycare increasingly rival the costs of rent—has since caught up. Now, neither party is budging. The study finds that “the impact of children on women is large and persistent across a wide range of labor market outcomes, while at the same time men are unaffected. The female child penalty in earnings is close to 20% in the long run.” Driving that wage penalty are “sharp impacts of children on labor force participation, hours worked, wage rates, occupation, sector, and firm choices.” While Denmark is not the most progressive among Nordic nations, no country has succeeded in completely closing the gender pay gap.

Women and men make different choices about whether and how to use the amenities on offer because those amenities impose opportunity costs at work, and because the relative value of the amenities—as well as the opportunity costs of using them—differs, on average, between women and men. Gender norms partly explain this. These norms no longer tell us that women should be full-time mothers and men full-time workers. But they evidently do influence how we navigate tradeoffs at the margins: women can care about their careers, but not if that means they aren’t fully present for their children. For men, social norms ask that they be fully present for their kids only when they’re present. Career obligations come first.

Even if these norms have only a slight effect—and their effect is arguably more than slight—they can generate powerful path dependencies and economic incentives. The parent who takes time off with her child will have lower monetary opportunity costs to cutting back later, because she’ll have missed an annual raise or opted for a more flexible position. And when a child needs a parent, it is always tempting to think she needs the most familiar parent. For these reasons and more, women are likelier to take flexible scheduling options and parental leaves. Because such options are negatively priced in labor markets, nominally gender-neutral family support policy and work time regulation won’t effectively combat gender inequality in caregiving and the workplace. Policies that ease women’s temporary exit from the labor force or their entry into more flexible tracks within it are an imperfect solution at best.

---

To combat gender inequality, family support policy must offset the influence of gender norms. It must nudge men to do more childcare and women to forge on at work. Over time, it must erode gender norms by incentivizing behavior that flouts them: fathers and mothers both utilizing family-friendly provisions to be actively and comparably involved in the work of caring for children. 

Such policy might involve:

• Packaging relatively short leave provisions with relatively generous daycare subsidies to encourage more families to opt for substitute care rather than for mothers to spend extended periods of time out of the labor market.
• Investing less in caregiver resource accounts or cash subsidies, which lessen the financial burden of women’s withdrawal from work, while investing relatively more in temporary leaves and subsidies for substitute care.
• Paying high wage-replacement rates to lessen the disincentive for high earners to take leave. It is tempting to contain costs by capping pay for leave-takers below 100 percent of their usual earnings. But partial wage-replacement can discourage leave-taking by the higher earner in a two-job household. Since high earners are still disproportionally men, discerning the threshold below which partial wage replacement constrains uptake of leave and keeping wage replacement rates above that threshold can encourage paternal leave-taking. (High wage replacement rates should be funded through strongly progressive revenue generation mechanisms.)
• Combatting workplace cultures that discourage male leave-taking, for example by imposing penalties on workplaces that fall below a set rate of leaves taken by eligible fathers.
• Attaching conditions to leave entitlements to encourage leave-taking by men, such as portions of leave available only to fathers. (In 1997 Denmark established two weeks of leave available only to fathers, but abolished that leave in 2002, despite the popularity of similar policies in other Nordic countries and their demonstrated efficacy in prompting fathers to take more parental leave.)
• Encouraging some leave to be taken very early on by the non-birthing parent: for example, structuring parental leaves such that each parent gets an additional month of leave on the condition that the non-birthing parent takes two months of leave immediately after the birthing parent’s medical leave ends.

We also need policies to induce workplace culture reform and increase flexibility and time off, of course. But when it comes to the details of caregiver support policy, we should not aim for maximal flexibility. There is evidence enough for optimism that policy details such as these could promote more active fathering and more gender-equal parenting in the short term, thereby lessening gender inequality in earnings and, in time, eroding gendered parenting norms.

---

Should the government really be in the business of intentionally influencing families’ work configurations and reforming social norms? 

The policy details proposed above put the legitimacy worry into sharp relief. Intrusive and inflexible family support policy looks like illiberal family support policy. It looks like a call to use political power to promote a contested set of values, at a cost to citizens who reject those values, on the grounds that the values are true or especially worthy of promotion. Many feminists think de-gendered parenting is an ideal worth striving for. But liberal feminists should tread carefully in endorsing social policy to realize that putative ideal, just as gender traditionalists should tread carefully in their calls for social policy to realize their ideal of the good society. How can such intrusive social policy, evidently justified on the basis of contested feminist values, constitute a legitimate exercise of political power in a liberal society?

Behind these worries stands a commitment to state neutrality: the idea that the state should not act on the basis of a contested ideal of the good life. Neutrality does not forbid social policy that disproportionally burdens or taxes certain values. Any social policy—and any omission of social policy—will impose disparate costs and benefits on different citizens depending on what they value and how they want to live. Consider an example: public funding for compulsory education can plausibly be said to impose higher monetary costs on citizens without children, since they pay into an education system that benefits everyone but benefits those with children most. And the compulsory part of public education plausibly imposes higher social costs on citizens who participate in certain tradition-based ways of life, since compulsory education induces the kind of exposure to questioning and diversity that is a driver of defection, thereby making such ways of life difficult to sustain over time. These disparate costs may be perfectly compliant with neutrality, depending on the reasons we can invoke to justify them. 

State neutrality proscribes all reasonably contested value commitments about what constitutes a good life as inadmissible grounds on which to justify public policy. Whatever disparate impacts they will invariably have, policies must be justifiable without invoking such contested values. Of course, public policy must be informed by some values. The values we may invoke to justify social policy include values that are part of the public political culture: values that enjoy broad support from a diverse array of citizens. So, for example, we may not institute a strongly progressive inheritance tax on the grounds that the interests of the rich are of comparatively little importance or that an acquisitive worldview is fundamentally flawed; but arguably we may institute such a policy on the grounds that it is necessary to ensure broad equality of opportunity in competitions for better-rewarded social positions. This is because—just assume for illustration if you’re not convinced of this—antagonism toward the wealthy is not a value that enjoys broad support from a diverse array of citizens, but equal opportunity is.

Flexible family support policy can be justified on the back of values that enjoy broad support from diverse citizens who have different ideas about what makes for a good life or a good society. We might make the case, for example, that well-raised children generate public benefit and so the public should support those who do the raising. But to argue for family support policy that is designed explicitly to erode gendered parenting norms, we need to argue for the erosion of gendered parenting norms as a policy end in its own right. 

Feminists who share my vision of a society free of gendered parenting norms can, consistent with liberal neutrality, endorse social policy to bring that ideal about. More importantly, precisely because such policy can be endorsed on neutral grounds, such feminists can insist that even those who don’t share their vision of the good society should share their commitment to eroding gendered parenting norms.

How?

---

As we’ve seen, neutrality forbids invoking certain contested values to justify social policy, but it permits broadly shared values to do that work. More strongly still, it designates certain political values as values that we can take for granted as shared, because those values encode fundamental interests we all share as free and equal citizens. These shared values include the presumption that we ought to be treated as free and equal citizens, and that we have fundamental shared interests in protecting our status as free and equal citizens and in preserving mutual respect among citizens construed as such. We can take these fundamental interests to be common ground, despite our many disagreements. 

Why do these values have such a privileged status? One way to see the answer is to notice that those values are what justify a commitment to liberal toleration—to neutrality—in the first place. We should abide by neutrality precisely in order to preserve mutual respect among citizens. In a liberal democracy, we construe ourselves and all other citizens as political equals with a presumption of freedom to pursue the good life as each sees it. To enjoy the benefits of living in a cooperative society with other citizens, we all accept some constraint on that prerogative. Neutrality tells us that the constraints we face are legitimate if and because they can be justified to us on terms that we can accept in our capacity as free and equal citizens. By acting democratically only in mutually justifiable ways, we express civic respect even to those who are constrained or burdened by our so acting. This means that we can’t invoke reasons that are exclusive to some disputed view of the good and that we can’t fail to heed those reasons that are implied by the liberal democratic values at the theoretical bottom of the whole enterprise. Social policy supported by the bedrock commitment to mutual respect among free and equal citizens is social policy that we cannot legitimately fail to enact. 

We can see, now, that a commitment to liberal neutrality does not entail a commitment to keeping all values out of politics. It implies a commitment to keeping out of politics those values that are not either broadly implicit in our public political culture or implied by the fundamental commitments that justify liberal neutrality in the first place. In short, the ideal of liberal legitimacy cuts both both ways: if policies to de-gender parenting can be justified only by invoking contested feminist values, then those policies are illegitimate. But if such policies are favored by the fundamental shared interest in preserving mutual respect among free and equal citizens, then it is illegitimate not to enact them.

The question thus becomes: Is there a case to be made on the grounds of mutual respect in favor of intrusive, inflexible family leave?

---

First, let us consider one way that case can’t be made. We might try to argue for social policy to de-gender parenting by invoking the distributive injustices that gendered parenting perpetuates. Women’s greater share of caregiving leaves them materially worse off than men and disproportionally materially poor in absolute terms. We can argue that poverty is a social ill without invoking any reasonably disputed premises; indeed, that conclusion easily flows from the fundamental shared value of mutual respect among free and equal citizens. We clearly fail to show mutual respect when we allow some among us to live in poverty. Mutual respect plausibly does not impugn all material inequality; a line will need to be drawn. But if some amount of distributive equality is a shared value, and if gendered parenting causes distributive inequality that the shared value impugns, then perhaps we can justify intrusive, inflexible family support as legitimate on those grounds.

The problem is that such support is not necessary to realizing distributive equality. Even if de-gendering parenting would help to ameliorate poverty or bring about greater distributive equality, those distributive aims can be accomplished by other, less intrusive means, too—such as labor market reform, taxation policy, and provision of in-kind goods. This is true whether we think of the goal as distributive equality between women and men, or between parents and non-parents, or between primary parents and secondary parents. Setting aside the question of whether any of these forms of equality is favored by a shareable ideal of distributive justice, all could be achieved without de-gendering parenting. The upshot is that those who think we have reason to de-gender parenting cannot justify policy to accomplish it on the back of distributive justice.

I think that distributive equality in some form is a demand of mutual respect among free and equal citizens. But distributive equality is not what the “equal” in “free and equal citizens” means. Equal citizenship refers to a kind of social equality: to relationships of equal respect, equal civic standing, and non-domination. It is because we are presumed to be equals in this sense that mutual respect demands for each of us a justification for political action that we can accept: one based on reasons that we can recognize as such. 

Can equal citizenship justify social policy to de-gender parenting? Not in any very straightforward way. Those who do relatively more parenting and who participate relatively less in labor markets plausibly do not currently stand as social equals. But the problem here is that caregiving is so devalued that it is not regarded as a source of social status in its own right. We should restore social equality by elevating the status of caregiving, and we can do this without persuading women to do less parenting or persuading men to do more. As someone who favors de-gendering parenting, I think we should work to equalize shares of it. But this cannot be for the sake of equal standing, because equal standing surely gives us more reason to elevate the status of roles that are currently unfairly and mistakenly devalued than to see to it that those devalued roles are shared equally between women and men.

---

If distributive justice and social equality cannot justify intrusive family support policy, what can? The missing ingredient is the free in “free and equal citizens.” 

To see why distributive fairness and social equality are not enough, suppose we achieve both: suppose caregiving and labor market participation are fairly rewarded and those who do them stand as social equals. Still, women might be socially steered into more caregiving-intensive roles and men into more labor-market-intensive roles. The source of this steering is the socially embedded assumption that sex dictates work specialization. And here’s our case for intrusive and inflexible family leave policy: that source of steering is objectionable from the perspective of free and equal citizenship.

Here’s how the argument goes:

The socially-embedded assumption that sex rightly dictates work specialization is objectionable from the perspective of free and equal citizenship because that assumption is inimical to autonomy. An autonomous person can reflect upon, revise, and reject the social roles and affiliations that fundamentally shape her life. Her choices in domains central to her identity are not effectively determined by normalized and institutionally-embedded assumptions that members of a social group to which she belongs are best suited to populating particular roles. We can be autonomous in a gendered society, because social norms do not determine our choices. And while gender norms attach social costs to some of the options we choose among, that, in itself, is no problem from the perspective of autonomy. Our options always carry contingent social and material costs. What’s objectionable is the particular configuration of social and material costs that gendered parenting norms sustain. Importantly, a social arrangement can be an affront to some value even without making realization of that value impossible. This can occur precisely when the arrangement is predicated on the assumption that citizens will not realize or aspire to realize the value in question. Our social arrangement is inimical to autonomy not because it makes autonomy impossible but because it presumes that citizens will behave non-autonomously: that they will specialize by sex into caregiving and labor market roles. Even without making gender-equal parenting impossible, and even without making autonomous choice impossible, our social arrangement is an affront to autonomy because it is predicated on the institutionalized assumption that one’s sex will dictate the work that one does.

How does that gendered assumption infuse our social arrangement? Labor markets are still designed for workers with “someone else at home” to specialize in parenting so that the “working” partner can devote himself fully to wage earning. Employers are permitted to impose demands on workers that are incompatible with workers having serious caregiving commitments, and success at work requires living up to such demands. The institutionalized assumption that sex determines work specialization explains why employers have been empowered to impose those demands. It also explains the dearth of support for substitute caregiving in the United States. Because we assume parents—mothers—will internalize the costs of caring for their children, we have neglected to develop social mechanisms for sharing those costs. And, as we’ve seen, the gendered assumption that explains this institutional arrangement continues to influence our strategies for navigating it.

The entrenched assumption that sex matters for parenting makes it socially and materially costly to avoid gendered parenting. A social arrangement that presumes that sex has this importance—and thus makes it very costly to arrange one’s domestic life in defiance of that presumption—is an affront to autonomy. And these institutional arrangements and social norms are objectionable on grounds of autonomy even though individuals can autonomously choose to comply with or flout those norms.

Gendered parenting is thus objectionable at the bar of mutual respect among free and equal citizens. Because that bedrock liberal value of mutual respect supports a case for de-gendering parenting, all of us committed to that value—and to the constraints and demands of liberal legitimacy that rest atop it—should demand policy to bring our society more fully into alignment with it. There is no illiberalism here, no illegitimacy. We have a compelling case for intrusive, inflexible family support policy that even those who care nothing about gender equality must accept.

 

Note: Adapted with permission from Liberalism, Neutrality, and the Gendered Division of Labor, Oxford University Press.