Introduction

The COVID-19 pandemic resulted in widespread school closures throughout much of the world. In this article, I argue that justice requires us to go beyond merely compensating children for educational losses stemming from such closures and instead aim for total equality in education. To do so, I’ll draw on Luck Egalitarianism, which is currently one of the most popular approaches to distributive justice in contemporary philosophy. The central idea of luck egalitarianism is that “it is unjust if some people are worse off than others through their bad luck” (Lippert-Rasumussen 2016, p.1). In other words, when inequalities are a matter of bad luck rather than individual choice, justice requires us to rectify those inequalities. Taking this central premise of luck egalitarianism—the claim that we should compensate people for unchosen inequalities—it seems simple and direct to apply luck egalitarianism to Covid-19-related school closures and conclude that we owe children compensation. After all, children did not have much choice in regard to school closures, and there’s substantial evidence to indicate that on the whole such closures have negatively impacted children.Footnote 1

In spite of this, I’ll argue that things are not as straightforward as they first appear. This is because it is not actually clear if and how luck egalitarianism applies to children. The theory must first deal with the agency assumption, an enduring problem for most theories of justice, which I believe to be particularly troubling for luck egalitarianism. The agency assumption highlights how theories of justice often assume a substantial account of agency which is not representative of children’s actual agential capacities. Recognizing the agency assumption yields what I call the deflationary dilemma for luck egalitarianism: either luck egalitarianism does not apply to children, or it does apply to children, but then it is no longer luck egalitarianism!

However, I believe this dilemma is surmountable if we recognize the implications of children’s actual agential capacities for luck egalitarianism. When we do so, we see that what looked like a fatal horn of the dilemma actually tells us something interesting about what luck egalitarianism requires when compensating children for school closures: justice requires that we aim to eliminate or neutralize all inequalities of currency between them, COVID-19-related or otherwise. Thus, although calls for post-Covid-19 educational responses have often focused narrowly on the need to compensate children for achievement loss, when properly understood luck egalitarianism leads us to a much more radical and universal approach: more than merely compensation for achievement loss, we need an explicit focus on achieving complete educational equality among children. This is a strong claim, and perhaps a surprising one from luck egalitarianism. However, this kind of radical equality is the only defensible conclusion from a luck egalitarian view when the agency assumption is taken into account.

Before continuing, allow me to clarify the scope of this article. First, both “children” and “childhood” are ambiguous and contested concepts, and it is difficult to draw a clear line between children, teeenagers, and adults.Footnote 2 Although these are important issues, I bracket them out by limiting my discussion to children under the age of twelve.

Second, I also set aside the question of what the proper currency of educational justice should be when discussing children. The question of currency is a serious question in itself, and one that I cannot adequately cover here. Fortunately, luck egalitarianism is typically understood to be neutral with regard to the proper currency of justice. Further, I don’t believe I need to specifically address the question of the currency of educational justice, as the issues I discuss in this article primarily stem from formal elements internal to luck egalitarianism as a theory, rather than substantive elements derived from its use of any particular currency. Instead, I believe it suffices to recognize that education (and the goods it contains) is likely irreducibly plural and complex. When I talk about (in)equality in this article, I am talking about (in)equality of education, where education is understood as a complex and plural term, a set of a wide variety of both instrumentally and intrinsically valuable goods.Footnote 3 Nothing in this article turns on what exactly this basket of valuable educational goods contains, so long as we recognize that the development of agency is certainly included.

Third, I’ll assume that whatever the proper currency is, it is provided for (in part, though not entirely) through school systems. I focus primarily on what schools and educational systems should do to address school closure related inequalities, rather than focusing on what parents and children can do. Although COVID-19 has affected almost every aspect of our lives in some way, and ameliorating COVID-19-related inequalities requires far more than educational institutions alone can provide, I limit my focus in this article to questions of justice in education. Although educational institutions do not provide the full picture, and other mechanisms for reducing COVID-19-related inequalities require consideration and discussion, I don’t believe I can adequately address all of these considerations within the scope of this article. Instead, I have set them aside for future discussion, and limited my discussion here to considering the question of COVID-19-related compensation solely in terms of education, and solely in terms of what can be done through educational institutions.

Fourth, when I discuss “inequalities” in this article, I’m only referring to justice-relevant inequalities. Not all inequalities are relevant to justice. For example, we are unlikely to think that inequality in the number of blue and green colored pencils each student has is relevant to justice. In contrast, inequalities in instruction time are much more likely to be considered justice-relevant. As I’ve left the currency of educational justice largely undefined—instead preferring to use the catch-all term “education”—I cannot offer any more specific discussion here nor can I offer an exact list of justice (ir)relevant inequalities. The question of which inequalities are irrelevant to justice is a point of major contention, and is beyond the scope of this article. In all, these four assumptions are not substantial, but they simply allow me to focus on my actual concern in this article: the adequacy of luck egalitarianism to determine the proper pattern of distribution for children.

Finally, allow me to clarify the relationship between the luck egalitarian dilemma I will discuss in this article and the COVID-19 pandemic. The dilemma children pose for luck egalitarianism is independent of the pandemic, and could have been raised in a non-pandemic context. However, I believe it is important to consider this dilemma in light of the pandemic. The discussion in this article is not just about the importance of this dilemma for COVID-19 policy; it is also about the questions such a pandemic raises for one of our “best” theories of justice. If one of the most popular theories of justice is incapable of guiding our response to this pandemic in a suitable way, then this is very bad news for the plausibility of the theory. This is an example of a ‘reflective equilibrium’ inspired approach, where the question raised is not just “what does this theory of justice tell us to do in this case?” but also “what does this case tell us about our theory of justice?” In this article, I’ll argue that the pandemic tells us that for the theory to successfully include children, it must be far more radical than it is typically assumed to be.

This dilemma is also important to COVID-19 policy if we believe our moral theory should have relevance for our policy. If we think that our theory of justice should inform our public policy, then it really does matter if it turns out that the theory does not apply to children. It matters both for how we assess our theory (as noted above) but if we are committed to the theory, then it also matters for what we should do in policy terms. If children really are not subjects of justice, then it doesn’t matter if we compensate them or not. Similarly, it matters to COVID-19 policy that the disadvantages among children are, as I will argue, incapable of being the product of choice. It is not uncommon for people to believe that education-related disadvantages are a product of merit. If we accept a view like this, when paired with an uncritical and unreflecting understanding of luck egalitarianism, we may not be so troubled by the rising educational inequalities arising from COVID-19. We may think that many inequalities result from differences in choices children made during periods of school closure and virtual schooling, and thus do not rise to the level of injustice. We may, if we accept such a view, believe that compensation must actually be significantly smaller, and importantly, significantly less egalitarian. The argument I make later in this article shows this view to be incorrect: children’s choices are not capable of justifying such serious inequalities. In short, although this dilemma can be discussed outside the context of the pandemic, I believe it is more informative to discuss them together, and that is the aim of the remainder of this article.

Do the Costs of Legitimate Public Health Measures Require Compensation?

In this section I do three things. First, I introduce the question of compensation for the costs of public health measures designed to control the spread of COVID-19, focusing specifically on the case of school closures. I then justify taking a justice-based approach to answering the question of how to compensate for school closures. Finally, I discuss the costs school closures have imposed on children, with a particular focus on their potentially lifelong effects.

To be clear, I will not address the question of the legitimacy of school closures as a measure to control the spread of the coronavirus. I set that question aside, and instead ask, assuming that such closures were and are legitimate, are children owed compensation as a result of these closures, and if so, what is the proper pattern of distribution? I also set aside other “knock-on” effects of school closures: e.g. family members being forced to leave their jobs to care for children and oversee their distance learning. I’m focusing narrowly on the question of what we owe to children specifically, as they have been often neglected in discussions of compensation for public health measures.

Despite the fact that some countries have already established laws or policies designed to compensate individuals for losses incurred as a result of public health intervention, philosophical and bioethical discussion of compensation for public health measures is limited. A notable exception is Soren Holm’s (2020) article, which offers a comprehensive and clear discussion of compensation in the context of reducing infectious disease spread. Holm aims to develop a general framework for determining compensation in cases where justified measures to limit the spread of infectious diseases impose costs on individuals.

Although Holm 2020 offers an excellent overview of compensation for public health interventions, he does not address school closures, nor does he specifically address children. However, school closures share important characteristics with the other disease control measures Holm discusses. Like the other measures discussed, school closures contribute to an important public health goal (limiting the spread of Covid-19), and they also impose significant costs on many—particularly on children. Further, given that school closures are typically state-led, such cases thus also entail questions about what society owes to those negatively impacted. As such, school closures, like other more widely discussed measures, raise the question of compensation.

Throughout the pandemic, there has been significant discussion of the economic and pragmatic reasons to compensate children for predicted achievement loss. However, I want to set aside economic and pragmatic reasons to compensate children for expected achievement loss, and instead examine the demands of justice in this area.Footnote 4

Why take a justice-based perspective? School closures’ negative impacts on children are particularly significant from the perspective of justice for at least two reasons. First, unlike other control measures discussed by Holm, and unlike other previously implemented public health related school closures (e.g. for scarlet fever or polio), Covid-19-related school closures offer little to no direct benefit for the majority of children. This makes the Covid-19 pandemic unique among public health directed school closures: past school closures resulting from pandemics have all involved diseases which pose significant risk to children. Previous school closures were justified (at least in part) on the basis that they constituted a direct health benefit to the children kept home. The same cannot be said of Covid-19-related school closures (Silverman, Sibbald, and Stranges 2020).Footnote 5

It is important to note that COVID-19 restrictions may have had some significant benefits for children. For example, reducing the spread of COVID-19 is important for protecting the health of adults who play central roles in children’s lives (parents, grandparents, teachers etc.). However, from the perspective of the direct health-related costs and benefits to children, school closures are unique. As noted, this is due to the fact that school closures were instituted to control the spread of a disease that itself does not constitute a significant threat to children’s health. This lack of threat is shown by data from UNICEF, which indicates no significant excess mortality among children during 2020, and from the ECDC, which indicates that child-to-child transmission in schools is uncommon (UNICEF IGME 2021, ECDC 2020).Footnote 6 Nor is it to reject the significance of the harm to those children, but simply to state that overall, it seems unlikely that children are at particular health risk from COVID-19, and this makes COVID-19-related school closures a particular concern of justice because school closures impose a cost on children primarily, though not exclusively, for the benefit of others. This asymmetry in the distribution of costs and benefits raises unique questions of justice.

Second, school closures are particularly relevant from a justice-based perspective because of children’s inherent vulnerability. Children’s vulnerability is complex and multifaceted, but the most salient element in this context is the fact that children’s vulnerability means that inequalities and disadvantages in childhood are likely to follow them for life.Footnote 7 As such, although there may also be other reasons to compensate children, exploring the question specifically from the perspective of distributive justice is particularly relevant and fills an important gap in the current discussion. So far, I have explained that Covid-19-related school closures raise questions of compensation. I’ve also suggested that a justice-based approach is particularly relevant for answering the question of compensating children.

To make the case that children are owed compensation as a matter of justice, however, two further things still need to be established. First, that children have been negatively impacted by Covid-19 related school closures, and second, that such negative impacts are unjust. In the remainder of this section, I briefly outline the negative impacts that school closures have had on children. Then, in the next section, I introduce the theory of luck egalitarianism, which will be used to determine whether the costs imposed on children by school closures are unjust.

Here, I briefly explain how children have been, on the whole, negatively impacted by school closures, particularly if we take a broad view of the importance of schools. My aim here is not to provide a comprehensive discussion of the negative impacts that school closures have had on children, but to offer a brief overview that highlights how complex and widespread such impacts are. Although there is currently some philosophical work discussing Covid-19-related intergenerational justice between the young and the old (e.g. Yarrow 2021), such work has so far not specifically considered the predicament of children in the pandemic. This is a major oversight, however. According to UNESCO data, in 2020 school closures affected around 1.5 billion students globally, accounting for 90 percent of students worldwide (UNESCO 2020a, as cited by Brando and Fragoso 2021). As noted earlier, there has been significant discussion about the need to ensure that predicted losses in educational achievement are compensated, but comparatively less work focusing on broader questions about compensating children for losses in education understood in a broader or more holistic sense. This is a myopic view, however, as a focus solely on learning achievement and educational outcomes distorts the true cost of school closures.

Schools provide more than achievement metrics, and school closures have cost children more than measures of educational achievement. Schools also play a significant social role in the lives of children. They are a place where children can socialize (both with other children and with adults outside of their family), as well as institutions that ensure that children are cared for in ways they may not be at home.Footnote 8 As such, not only have school closures resulted in a predicted achievement loss, but they have also left many children bereft of much needed care and support. In addition to such direct and short term impacts on children, there are also longer term effects, in part due to children’s vulnerability. These effects are particularly relevant from the perspective of justice. Without intervention, school closures are also likely to have broader social consequences, in that they will increase more general (educational, social, and economic) inequalities between and among children that are likely to extend into adulthood (Eyles et al. 2020; Hanushek and Woessmann 2020).

There is much more that could be said about the effects both the pandemic and its control measures have had on children, and it’s important to recognize that children were not uniformly or equally affected by pandemic measures, both across and within countries. It is worth noting that many students may have benefitted in various ways from school closures, e.g. either by avoiding schools with serious issues, or by developing greater resiliency in facing problems. However, I set aside the question of how to address the question of possible positive consequences for students. I do so for two reasons. First, as I discuss in this section, even if some students have benefited from school closures, on the whole school closures seem to have negatively impacted children. Importantly, the positive and negative impacts of school closures are not equally distributed, and school closures are predicted to significantly increase inequalities among children. This is clear from looking at the data on how wealthier and poorer students weathered school closures: wealthier students managed pandemic-related disruptions to their education better than their poorer counterparts.Footnote 9 Second, as I will discuss in the next section, luck egalitarianism is typically concerned about the negative differential effects of brute luck. In other words, some students benefiting is not necessarily a concern of (luck egalitarian) justice. I will address this further in the next section.

Therefore, for the sake of simplicity, in the rest of this article I will paint with a broad brush, and assume that all children suffered to some extent from school closures. Now that it has been established that children have been negatively impacted by school closures, the next question that must be answered from this justice based perspective is whether or not such negative changes are unjust. If so, there is a justice-based obligation to compensate for such changes. In the next section, I introduce the theory of luck egalitarianism, which I will use to determine whether the negative impacts Covid-19-related school closures have had on children are unjust.

Luck Egalitarianism and Compensation for Unjust Inequalities

In this section, I offer a brief introduction to luck egalitarianism, and argue that applying luck egalitarianism to the case of school closures under COVID-19 suggests that children are not responsible for the inequalities stemming from COVID-19 related school closures, and thus justice requires that we compensate them.

I focus on luck egalitarianism for a few reasons. First, it is one of the most popular approaches to distributive justice in contemporary philosophy. Second, despite its popularity, luck egalitarianism has, unfortunately, rarely been applied to the question of children. Third, luck egalitarianism is particularly well-suited to address losses like those from Covid-19 control measures: it is not typically concerned with ameliorating every instance of inequality, but is instead is focused on inequalities stemming specifically from losses or negative changes.

Finally, the luck egalitarian approach examined here presents a major advantage over other approaches to determining compensation for children in the case of Covid-19-related school closures: it is comparatively very simple. There is no need to answer a variety of difficult questions that other approaches to compensation face. For example, luck egalitarianism does not need to answer questions about cost/benefit ratios, the justifiability of school closures, whether school closures have wronged children, and so on. Instead, as I’ll argue later, all that is required is recognizing that in a luck egalitarian framework, inequalities among children are solely the product of “brute luck,” and as such, children are entitled to compensation to neutralize the differential effects of such brute luck.

Although I’ll refer to luck egalitarianism as if it were a single unified theory, it is perhaps more accurate to characterize it as a family of closely related theories.Footnote 10 The fundamental concern of luck egalitarianism is to provide an egalitarian theory that is appropriately sensitive to individual responsibility. It is this sensitivity to individual responsibility that most clearly distinguishes luck egalitarianism from simpler forms of egalitarianism. Although luck egalitarianism is an egalitarian theory, it is not committed to condemning each and every instance of inequality. Instead, it makes a distinction between acceptable and unacceptable inequalities, a distinction premised on the idea that involuntary inequalities are unacceptable, whereas voluntary inequalities are acceptable. Arneson (2000 p.339) captures this core idea well: luck egalitarianism aims “to eliminate so far as possible the impact on people’s lives of bad luck that falls on them through no fault or choice of their own.” Luck egalitarians often use the terms “brute luck” and “option luck” to denote this distinction between acceptable and unacceptable inequalities. Inequalities stemming from an individual’s choices (“option luck”) are acceptable and don’t require compensation or redress. In contrast, inequalities that an individual is not responsible for (“brute luck”) do require compensation.Footnote 11

The main take away from this brief introduction to the theory is that every luck egalitarian theory requires inequalities to be the product of either option luck or brute luck. The type of luck that results in a given instance of inequality determines whether or not the inequality is required to be rectified as a matter of justice. Inequalities caused by option luck do not, as a matter of justice, require compensation, whereas inequalities from brute luck should be compensated. Applying this brief description of luck egalitarianism to the earlier discussion of how school closures will make children worse off and increase inequality yields the conclusion that children should be compensated for the inequalities stemming from Covid-19-related school closures.

This conclusion is not particularly surprising or innovative, and doesn’t really require any particularly interesting philosophical argumentation. Instead, it’s a pretty straightforward consequence of most (if not all) formulations of luck egalitarianism.Footnote 12 Although this conclusion seems straightforward, it obscures a deeper and more difficult question: can luck egalitarianism actually be applied to children? I believe that a common problem called the agency assumption raises a difficult dilemma for luck egalitarianism. In the next section, I examine this assumption and argue that the dilemma it presents is escapable if we recognize a couple key features of children’s reduced agency.

The Agency Assumption: Does Luck Egalitarianism Apply to Children?

Although the conclusion that we must compensate children for inequalities resulting from school closures seems straightforward, things are actually not so simple. When applied to children, luck egalitarianism faces a problem shared by many other theories of justice, often called the agency assumption. In this section, I explain the agency assumption, and then explain how it is a particularly serious problem for luck egalitarianism. I argue that recognizing the agency assumption results in an interesting dilemma for the theory, which I call the deflationary dilemma. I claim that this dilemma is surmountable, and that addressing it reveals something interesting about compensating children for Covid-19 related school closures: in a luck egalitarian framework, there can be no acceptable justice-relevant inequalities among children.

Macleod (2010) explains the agency assumption in his critical discussion of both Rawls’ theory of justice and the capability approach. The agency assumption is, “a common implicit assumption … that the human beings with whom justice is concerned are to be conceived primarily either as mature agents capable of assuming responsibility for the ends they adopt and pursue or as potential mature agents” (Macleod 2010 p. 174). However, as Macleod argues, “children do not comfortably fit the paradigm of agency” that is required by this assumption (Macleod 2010 p. 174). he core of this assumption is that persons are agents capable of exercising responsibility, acting responsibly, and taking responsibility for the ends they set. Although Macleod focuses on Rawls’ theory of justice and the capabilities approach, the problem he raises actually applies in a general form to many theories of justice (Lindblom 2016, p.98). Although Macleod 2010 (and later Lindblom 2016) both focus on currency, the same problem actually applies to the pattern of justice, at least in the luck egalitarian case.

Like most theories of justice, luck egalitarianism struggles to address the agency assumption. However, the agency assumption proves to be a particularly significant challenge to luck egalitarianism specifically. To see how, we need to ask about why luck egalitarianism holds option luck to matter in such an important way. In other words, the important question here is why holding people responsible for their choices matters, in a moral sense. This question can be answered by an appeal to the value of agency. We might have certain pragmatic or practical reasons to hold people responsible for their option luck, but this does not explain the moral importance that luck egalitarianism ascribes to the distinction between option luck and brute luck. Ultimately, what is important about option luck is that it is a reflection of one’s agency. As Axelsen and Nielsen (2020) explain,

respecting people’s choices is important because such choices are a product of moral agency (Arneson 1999; Dworkin 2000, chap. 1; Lippert-Rasmussen 2016, chap. 2). Luck, on the other hand, does not originate in moral agency and, thus, should not be treated with similar respect (Otsuka 2002; Stemplowska 2008). With this grounding in mind, it becomes clear why distributions of benefits and burdens should reflect people’s choices, but not natural and social contingencies. (pp. 657-658. Citations included in original quotation.)

By and large, luck egalitarian theories have not paid much attention to the question of agency. It has simply been assumed that everyone has the right kind and degree of agency. However, as Macleod argues, this is a flawed assumption in the case of children. Children do have agency to a considerable degree, and as has been pointed out, children’s agency is often incorrectly assumed to be far weaker than it actually is (Cf. Gheaus (2015), especially pp.16–17). However, although children may have agency, their agency is limited in various important ways, and as such they lack the kind and degree of moral agency that luck egalitarianism (as well as other theories of justice) assumes.Footnote 13

In short, the problem posed by the agency assumption is that differential outcomes from option luck are supposed to reflect a respect for people’s agency. Children, as the above discussion indicates, typically do not have the same kind and degree of agential capacities that luck egalitarianism assumes.Footnote 14 Since children lack the requisite capacities that option luck requires, no actions carried out by children can be counted as option luck.Footnote 15 So, option luck—that is, holding children responsible for their choices—cannot justify inequalities in currency. As a result, all inequalities between children should be considered products of brute luck and compensated as a matter of justice.

The Deflationary Dilemma for Luck Egalitarianism

This lack of agency (and its accompanying lack of option luck) results in what I call the deflationary dilemma. Recognizing the problem that the agency assumption poses for luck egalitarianism threatens to impale the theory on one of two horns of a dilemma: either luck egalitarianism does not apply to children, or it does, but then it is no longer luck egalitarianism at all! This dilemma stems from the fact that, as discussed earlier, children do not have option luck. In non-luck egalitarian parlance, this simply means that children have reduced moral responsibility, and so it does not make sense to hold them accountable for their choices in the same way as it would adults. This is not to say children are never responsible for their choices or anything similar, but simply that within the definitions of luck egalitarianism, children do not have the kind of responsibility that the theory requires for option luck to be a legitimate (i.e. not unjust) source of inequality.

This lack of option luck leaves the luck egalitarian with two choices, representing the two horns of the dilemma. First, she could deny that luck egalitarianism applies to children (or, even more radically, she could deny that theories of justice apply to children at all). Under this view, the seemingly straightforward conclusion from earlier is actually incorrect: a luck egalitarian theory has nothing to say about whether we owe children compensation for school closures. Alternatively, she could accept that children do not have option luck, and that they cannot be held responsible for inequalities in currency. In other words, this means holding that every instance of justice-relevant inequality between children is simply a matter of brute luck and thus must be compensated.

This second option is similar to the “all luck” view of luck egalitarianism. The all luck view is a particularly extreme version of luck egalitarianism that holds that people are almost never responsible for their choices.Footnote 16 At its core, the all luck view holds that “differential option luck should be considered as unjust as differential brute luck” (Segall 2010 p. 47). However, the all luck view faces a rather devastating objection. It presents a radically deflationary view of luck egalitarianism because in many cases (perhaps even all cases) it is extensionally equivalent to simple egalitarianism (Hirose 2014 pp.51–54). In other words, the problem is that the all luck view is not really luck egalitarian because it seems to make little to no use of the notion of the distinctive and defining element of the theory, namely holding people responsible for their option luck. This is one of the strongest objections to the all luck view, and it seems to apply particularly forcefully to the account I’ve presented here. If children have no option luck, then how can this theory still be called luck egalitarianism? The defining feature of the theory—that inequalities caused by option luck are acceptable—is absent. Hence, the dilemma:

The Deflationary Dilemma Either luck egalitarianism does not apply to children, or it does apply to children, but then it is no longer luck egalitarianism.

Escaping the Dilemma: a Nuanced View of Luck Egalitarianism

As a result of this dilemma, it may seem that luck egalitarianism is inapplicable to children. It may appear to be the case that luck egalitarianism is a theory designed only for adults, and is incomplete without an accompanying separate theory for children. I think this conclusion is too hasty, however. In fact, I believe that accepting one horn of the dilemma—accepting children’s lack of option luck—is not as troubling as it first appears. To see how this horn of the dilemma is not as fatal as it first appears, we need to recognize two interrelated features of children’s lack of responsibility: it is not permanent, and it is not global.

First, we should recognize that children’s lack of agency (and corresponding lack of option luck) is not permanent; as children age and mature, their agential capacities also develop. Thus, this lack of option luck is not permanent, nor is its accompanying lack of responsibility. So, while luck egalitarianism may not initially look much like luck egalitarianism when applied to children—because it does not hold them responsible for differences in the distribution of currency—I don’t believe this entails such a radically deflationary view as critics of the all luck approach would claim. Children’s lack of option luck is tied to their agential capacities. As their capacities increase, so does the amount for which they can be held responsible.

This first feature leads to the second: we should recognize that holding children responsible for some choices does not require holding them responsible for all choices. Children develop agency (and responsibility) as they age and mature, and yet, holding children responsible for their choices seems to be central to the successful development of their agency. As Drerup (2016) explains, “it is a pedagogical truism that children have to take responsibilities within appropriate constraints to become able to learn to understand themselves as autonomous agents who are responsible for their own actions” (Drerup 2016 p.126). It seems undeniable that children do have agency, and that they must exercise this agency and be held responsible (within certain limited constraints) in order to further develop and mature. This may seem to pose a problem for what I’ve claimed here, because it may seem that the demands of luck egalitarianism are contradictory—we must both not hold children responsible (because of their lack of agency) and hold them responsible (so they can develop agency).

However, I don’t believe this is a problem. Not only is children’s lack of responsibility not permanent, it is also not global: children can and should be held responsible for a variety of things—but not for inequalities in currency. This is because holding children responsible depends on more than simply their age and their agential capacities. It also turns on the importance of the decision at hand. A wide variety of important and potentially life altering decisions are not permitted to be made by children for precisely this reason. Generally speaking, the more important any given decision is, the less children should be considered responsible for it (Ben-Shahar 2016, p.92). Given the importance that any currency of educational justice holds for overall life success, it seems unjust to hold children fully responsible for inequalities in currency.

Taken together, I suggest that these two features indicate that the appropriate view of agency for luck egalitarianism is a nuanced one, which allows us to successfully embrace the “all luck” horn of the dilemma. As I noted in my discussion of the dilemma, I am not claiming that children are never responsible for their actions, or that they are completely without agency or anything similar. Rather, my claims are much more limited: children do not have the kind and degree of agency that makes holding them responsible for differences in the currency of justice a morally valuable activity within a luck egalitarian framework. Clearly, we can, and should, hold children responsible for a variety of choices and actions. However, this is not the same as saying that their choices should influence the overall distribution of currency—in other words, although we can and should hold children accountable in ways that help foster and develop their agency, this does not mean that their limited agency can justify inequalities in currency within a luck egalitarian framework.

Thus, the impact children’s choices can meaningfully have on the overall distribution of currency should vary in accordance with children’s agential capacities. It makes significantly more sense to hold older children more responsible than their younger counterparts. Not only is it more sensible, but it is also necessary for the development of children’s agency. Preventing childhood choices from influencing the overall distribution of currency is, in my view, a key part of ensuring and promoting the development of effective agency in children. As such any coherent luck egalitarian theory must also be committed to promoting these capabilities in children. This implies a gradualist view of option luck, which moves from the all luck view to a more traditional luck egalitarianism as children age into teenagers and then into adults.Footnote 17

This gradualist understanding of option luck emphasizes the importance of holding children responsible in ways that foster the development of their moral agency. It recognizes the fact that agency in children is a developmental process: we cannot simply declare them unaccountable for all choices, nor can we hold them accountable the way we do adults. Instead, we must recognize that holding children responsible for some of their choices is a necessary element of this developmental process. Such a view of children’s choices represents a “forward-looking” approach to children’s agency that focuses on helping children develop the kind of responsible agency necessary to be full-fledged moral agents in the way luck egalitarianism requires. This understanding emphasizes the importance of recognizing that children are not just defective moral agents, but are in a process of consistently increasing agential development, and thus, holding them responsible is a important practice that helps to inculcate the kind of agency we will expect from them as adults (cf. Burroughs 2020).Footnote 18

What Does Luck Egalitarianism Tell us About Compensating Children for School Closures?

Having now explained why luck egalitarianism applies to children, although not in the way one may expect, I return to the question of compensation for Covid-19-related school closures. I argue that luck egalitarianism—when properly understood in light of the agency assumption—requires us to go beyond a narrow view of compensation and instead aim for total equality in education.

The important take-away here is that although much of the current discussion of post-Covid-19 recovery strategies has focused on the need to ‘make up’ for educational losses stemming from Covid, when properly understood, luck egalitarianism actually requires us to do significantly more.Footnote 19 Luck egalitarianism requires us to address all instances of justice-relevant inequality among children—Covid-19-related and beyond.

As such, luck egalitarianism emphasizes that Covid-19-related school closures should be understood as worsening already existing educational injustices rather than as something entirely new. Although the pandemic may be unique in terms of the scale and speed at which it has increased inequality, the inequalities driven by school closures are not new nor limited to the pandemic. Prior to the pandemic, an OECD report released in 2018 revealed that, although both educational access and educational attainment have increased over the past half-century, educational equity has not followed the same trend. (OECD 2018 p.24). Further, the data collected by the OECD suggests that already advantaged students benefited more from this increase in access to education, a trend likely intensified by Covid-19-related school closures, as disadvantaged students suffer more than their advantaged counterparts (OECD 2018 p.24; Di Pietro et al. 2020 p.12; Eyles et al. 2020; Kuhfeld et al. 2020). Due to children’s vulnerability, these disadvantages may follow students for life, and result in significant social and economic inequalities if they are not mitigated. However, due to the agency assumption, these inequalities cannot be considered to be the result of option luck. What this means is that luck egalitarianism, as a matter of justice, requires us to go beyond a narrow focus on making up for achievement loss and instead aim to ameliorate all justice-relevant inequalities among children. Thus, this luck egalitarian view requires us to ensure that no child is worse off in terms of education solely because of luck. And as explained earlier, due to the way luck egalitarianism links luck and agency, (almost) everything is a matter of luck for children. Consequently, luck egalitarianism requires us to set up our educational institutions to neutralize and compensate for all instances of justice-relevant inequality in children’s education. This is a radical conclusion, but it is not an outright implausible one. There are already plausible arguments for, and defenses of, “all the way equality,” although none from a specifically luck egalitarian view so far.Footnote 20

Such an “all the way” equality view does encounter significant problems, but these problems are also common to egalitarian theories that address children. For example, there is significant tension between the assumed autonomy of family life and the goal of neutralizing inequality, and it is an open question as to how to appropriately balance them. However, this is a problem common to many egalitarian views, and is not unique to luck egalitarianism.Footnote 21 Although the luck egalitarian view discussed here faces problems common to other egalitarian views, it is also worth noting that the view discussed here may perform slightly better than some other egalitarian views in this regard. This is because the luck egalitarian view discussed here focuses on negative differential luck. This helps to limit some of the scope of justified intervention, e.g. it limits the extent to which we should be concerned about parents conferring advantage to their children.Footnote 22 It’s also worth noting that I am not defending luck egalitarianism as the best theory of justice, nor am I arguing that it has no counterintuitive conclusions. My aims in this paper are much more limited: I aim to discover what luck egalitarianism tells us about compensating children COVID-19-related school closures, and what this response to COVID-19 tells us about the adequacy of luck egalitarianism. Examining this response has revealed that luck egalitarianism requires us to adopt a much more radical understanding of egalitarianism than typically assumed.

Conclusion: Luck Egalitarianism, School Closures, and Radical Equality

In short, luck egalitarianism fundamentally links the notion of (un)justified inequalities and agency. This leaves it vulnerable to the agency assumption, which highlights how theories of agency have traditionally neglected children’s actual capacities. When this assumption is taken into account, the result is a radical form of luck egalitarianism that advocates for complete equality among children. In the case of Covid-19 related school closures, this suggests two things: First, a need for luck egalitarianism to focus on the development of children’s agency; Second, and more importantly in the context of this article, in a luck egalitarian framework there can be no acceptable justice-relevant inequalities among children. So Covid-19, although exceptional, needs to be understood as worsening already existing inequalities, all of which must be addressed.

To address the first point, one consequence of this view of luck egalitarianism is that the theory requires us, as a matter of justice, to develop children’s agential capacities. Regardless of the actual equilisandum we select, the theory cannot function without agents with substantial agential capacities. The central defining feature of luck egalitarianism—that option luck can justify inequalities—is premised on such capacities. As such, the theory must emphasize the importance of developing children’s agency. Second, since within luck egalitarian theories the only acceptable deviations from equality stem from option luck, and children do not have option luck, there can be no acceptable deviations from equality for children. This leads to a surprisingly radical luck egalitarian view, that holds that the goal of justice should be total equality in the currency of educational justice for children. Some of exactly how radical this approach is depends on the currency of educational justice in particular. I should note that although the luck egalitarian approach developed here requires strict equality, this obviously may not always be possible. Children do have different capacities, and no matter how much we intervene we can never make the currency of justice entirely equal between children. While that may be true, the central point still holds: we are required to take significant measures to reduce inequalities between children.

Given the need to compensate children for inequalities stemming from school closures, there is still the question of when we should intervene to ameliorate these inequalities. Should we aim to ameliorate inequalities now, or wait to compensate them for disadvantages in the future.Footnote 23 This is particularly relevant because many of the inequalities driven by school closures are predicted to arise in the future, and are not currently visible, e.g. reductions in future income. Many of the proposed pathways for how COVID-19-related closures are likely to result in increased inequality—e.g. by increasing gaps among socio-economic groups in access to higher education—are likely to compound over time, and to further entrench intergenerational inequality. As such, addressing current inequalities among children seems to be a promising route to mitigating inequalities among adults in the future, in particular because all childhood inequalities are unjust, whereas some inequalities among adults are, according to luck egalitarianism, just. Consequently, it is significantly easier to intervene now to address inequalities, without having to determine which are just and which are unjust. Additionally, there is the practical consideration of political will. Given that the costs of compensating children for COVID-19 are likely to be very high, the further we get from the pandemic, the less salient rectifying the inequalities from it may seem. It seems prudent to attempt to implement compensation policies sooner rather than later, if only because it may make it easier to capture popular support for such policies. As such, in this article, I focus on what luck egalitarianism has to say about immediate interventions aimed at ameliorating COVID-19-related disadvantages. My focus is on what we owe to children now, as children, rather than what we may owe to them later, as adults.

There is a great deal more to discuss, particularly regarding exactly how luck egalitarianism should proceed with reducing inequality among children in general. This is beyond the scope of this article, however. Instead, I’ve focused more narrowly on the fact that when applied to children, luck egalitarianism must adopt a forward-looking and developmental approach which aims at (1) fostering the appropriate kind of agency in children that they will need to be luck egalitarian agents as adults, and most importantly (2) eliminating inequalities between them to the furthest extent possible—particularly and most importantly, those inequalities that affect the development of children’s agency. In the context of Covid-19 related school closures, this means that mere compensation for achievement loss is not enough: justice requires complete equality of education for children.