The Distributive Demands of Relational Egalitarianism

Abstract

The article outlines the distributive demands of relational equality in the form of a dynamic corridor of legitimate distributive inequality. It does so by complementing the already widely accepted sufficientarian floor with a limitarian ceiling, leading, in a first step, to a "corridor" of limited distributive inequality as a necessary condition for relational equality. This corridor alone, however, only provides necessary distributive conditions for relational equality and still allows for degrees of distributive inequality that would risk undermining egalitarian relations. Thus, in a second step, intra-corridor distributive inequalities must be regulated by two further constraints: a (context- dependent) ratio between the best- and worst-off, and a demand for equality of opportunity so that inequalities result from people’s responsible choices. This set of demands spells out the distributive conditions that are not only necessary, but sufficient to provide a sound distributive basis for relational equality. After presenting this view, the article defends it against several objections.

“So distribution should undo excess, And each man have enough.”

–– Shakespeare, King Lear.

1 Introduction

What, if any, are the distributive demands of relational egalitarianism? This important question arises with the distinction between distributive and relational accounts of egalitarian justice: If one stipulates relations (instead of distributions) to be the primary concern of egalitarian justice, does this imply that distributions do not matter at all? Few would answer this question in the positive. Standardly, relational egalitarians demand sufficiency with regard to the distribution of what is considered to be the relevant metric of justice, so that people can live and interact with one another as equals (Anderson 1999, 317–318). Others have argued relational egalitarianism would start off by implying some kind of equal distributions, a “presumption of distributive equality” (Schemmel 2021, p. 236) or e.g. “roughly equal levels of (weighted) interest satisfaction” (Scheffler 2015, p. 34). But sufficiency on the one hand may not be enough to provide a basis for egalitarian relations, as residual inequalities above the sufficiency threshold can negatively impact upon relations and interactions. On the other hand, a stricter demand for distributive equality may appear unnecessarily demanding, even as an ideal, since it is neither desirable nor necessary to secure relational equality. Thus, further explorations are necessary to specify the distributive demands of relational egalitarianism, spelling out the “roughly equal”-demand in greater detail. A convincing account of the distributive demands of relational egalitarianism will name and justify criteria for legitimate distributive inequality that remains compatible with and is conducive to egalitarian relations. The present paper proposes and discusses such criteria.

In the following, I provide a first full outline of the distributive demands of relational equality in the form of a dynamic corridor of legitimate distributive inequality. I do so by complementing the already widely accepted sufficientarian floor with a limitarian ceiling, leading, in a first step, to a “corridor” of limited distributive inequality as a necessary condition for relational equality (cf. Heilinger 2020, p. 67). This corridor alone, however, only provides necessary distributive conditions for relational equality and still allows for degrees of distributive inequality that would risk undermining egalitarian relations. Thus, in a second step, intra-corridor distributive inequalities must be regulated by two further constraints: a (context-dependent) ratio between the best- and worst-off, and a demand for equality of opportunity so that inequalities result from people’s responsible choices. This set of demands spells out the distributive conditions that are not only necessary, but sufficient to provide a sound distributive basis for relational equality.¹

The main part of the paper (Sects. 2 to 5) spells out this view. The last section discusses its implications and some objections against it in the light of making use of the theory under the non-ideal circumstances of our current world.

2 Distributive and Relational Egalitarianism

The recent scholarly debate about equality and justice is structured by two main poles, one distributive and the other relational. Distributive and relational accounts of equality and justice are distinguished by their different starting points: patterns of distributions of advantages of all kinds, such as resources, well-being or capabilities on the one hand (Arneson 1989; Dworkin 2000); or the quality of relations and interactions among individuals or within communities on the other (Young 1990b; Wolff 1998; Anderson 1999).

Relational egalitarianism demands positively that all do relate to and interact with one another “on a footing of equality” (Scheffler 2015), i.e. in a way that the equally important needs and interests of all are weighing equally in decisions; and it demands negatively that all forms of oppression of individuals and groups – e.g. in the form of domination, exclusion, marginalisation, exploitation (Young 1990a) – must end (cf. Anderson 1999, p. 313).

I define relational egalitarianism fundamentally in practical terms that demand people display a particular “ethos”, i.e. a set of attitudes, motives, dispositions and habits when interacting with one another (Cohen 2008; Heilinger 2021). It demands mutual acknowledgement of everyone’s fundamentally equal moral worth and a willingness to treat one another with respect. Also, all must be seen as worthy of participating on equal terms in decisions of joint interest (cf. Scheffler 2015). Concretely, relational equality can manifest itself in agents exercising an “egalitarian deliberative constraint”, meaning that every person’s “equally important interests—understood broadly to include the person’s needs, values, and preferences—should play an equally significant role in influencing decisions made within the context of the relationship” (Scheffler 2015, p. 25).

I argue that the differences between distributive and relational accounts of justice must not be overstated, but merit more philosophical exploration. I suggest to understand them as a family quarrel among egalitarians, and that ultimately both aspects, relations and distributions, matter for realising a just society (cf. Heilinger 2020, p. 115). In particular, realising relational equality also comes with some distributive demands (Schemmel 2011, 2021), which are explored in detail in the following. From a relational perspective, however, patterns of distributions do not have intrinsic value. The reason for which relational egalitarians should be concerned about distributions is their instrumental impact upon relations.

3 The Dual Threshold Theory (DTT). The Basic Formulation

In an initial attempt to spell out the distributive demands of relational equality, I propose and discuss a corridor of legitimate inequality confined by a sufficientarian floor of “having enough” and a limitarian ceiling of “not having too much”. Such distributive demands will certainly emerge from a joint deliberation about distributions among people who respect one another as moral equals and are willing to exercise deliberative constraint. While a demand for securing enough for all and limiting excess may seem rather uncontroversial and intuitive to many, there are also strong arguments to support it. Making these arguments explicit is important because in particular the demand of limiting excessive concentrations of income and wealth runs against the widely accepted pursuit of becoming, or admiration for being extremely affluent that shapes many contemporary societies.

One last clarification is required before defending the two components of the distributive corridor: When discussing distributions I will, in the present context, focus predominantly on money as the relevant metric of just distributions. Money is understood as an all-purpose resource that can be easily measured, compared and transformed into many other goods or advantages. Choosing money as the metric also responds to the massive, and empirically well-researched economic inequalities that shape our current world, invites easier operationalisation in policy, and gives room to individual choice in how to spend it.²

So equipped let us now move to the two threshold concepts constitutive of the first attempt to specify a corridor of legitimate inequality. I will introduce and justify each threshold by looking initially also its non-relational justification, even if the relational justification which follows will be stronger and more relevant from the distinctively relational perspective endorsed in this paper. Yet, it will not harm to acknowledge that also non-relational justifications can be provided for each threshold concept.

3.1 Sufficientarianism

Relational egalitarians, insofar as they care about distributions, tend to call for sufficiency. Anderson, for example, demands that all must be able to function as equal citizens and lists what is necessary and sufficient to this end (Anderson 1999, 317–318). While she focusses on capabilities (in itself a sufficientarian concept), not money, it is clear that securing capabilities will often require the distribution of goods or taking measures that can be framed in terms of money. Other relational egalitarians have more directly targeted that people have “enough” of the relevant goods or resources.³ Such a sufficientarian demand is highly plausible, since it prioritises the legitimate needs and interests of all without excessively demanding strict distributive equality.

The positive claim for sufficientarian distributions can be based on the (non-relational) importance for all to be free from duress and deprivation (Axelsen and Nielsen 2015). Having one’s basic needs met and having enough for being able to function as a human being – and subsequently also as participants in a system of cooperative production, and as citizens in a democratic state (cf. Anderson 1999) – is particularly important also from a relational perspective: being free from deprivation and being able to realise basic functionings is a necessary condition for engaging with others as moral equals. This then is a distinctively relational justification for the sufficientarian threshold.

However, next to the positive demand that all have enough of the relevant distributive good, sufficientarianism as a distributive theory also involves the “negative” claim that supra-threshold inequalities do not matter morally (Casal 2007). Sufficientarianism’s negative thesis thus allows for – potentially immense – supra-threshold inequalities, and it is easy to see how amassing and controlling much more resources than others will undermine relational equality (more on this in the following section on limitarianism). This makes sufficientarianism, from the perspective of relational equality, at best only a partial and incomplete account of distributive justice. But the challenge that sufficiency alone may not be enough for distributive justice has been felt also by sufficientarians themselves who may or may not endorse relational egalitarian ideals. Recent developments in sufficientarian theorising have attempted to address this challenge, e.g. by so called: “shift-sufficientarianism” where distributive claims above threshold have different (but not no) moral weight (Shields 2012); by “multi-threshold sufficientarianism” where the positive and the negative claim of sufficientarianism do not necessarily attach to the same distributive threshold (Huseby 2020); or by “weighted sufficientarianism” where multiple thresholds differentiate the moral weight of claims to distribution according to the location of the recipient in a distributive hierarchy (Timmer 2022).

What is common to these approaches is the shared insight that supra-threshold inequalities are very much of moral relevance so that a single sufficientarian threshold alone fails to offer a complete account of just distributions. Given this critique, a first attempt to contain supra-threshold inequalities thus consists in stipulating at least one additional threshold concept: an upper limit to the morally acceptable concentration of advantage/the relevant distributive goods.

3.2 Limitarianism

Recent debates on distributive justice have increasingly directed attention towards “having too much”, so called “limitarianism”. Limitarianism is not so much concerned with what people should have or are entitled to get, but with what people should not have. Concretely, it argues that “it is not morally permissible to have more resources than are needed to fully flourish in life.” (Robeyns 2017, p. 1). Limitarianism, according to Ingrid Robeyns, is a “partial account of justice” that can be combined with additional distributive demands (Robeyns 2022, p. 256; Robeyns ed 2023). While limitarianism, as I will show, can indeed be part of determining the distributive demands of relational egalitarianism, the demand for an upper limit to the accumulation of income and wealth can initially be justified in its own standing, without reference to any distinctively relational arguments.

Having too much might be in some way intrinsically bad for those who do. This somewhat paternalistic reasoning is not central in Robeyn’s writings, but can still offer a first – although admittedly not the strongest possible – line of defending limitarianism: Having too much may be bad insofar as having too much – e.g. money, or opportunities, or possibly also well-being – might ultimately impact negatively upon one’s life: the abundance of available resources or opportunities might make choice difficult, distract people from what is genuinely valuable, or invite immodest, unhealthy, or wasteful behaviour. Concern for the fate of the super-privileged shows that some arguments speak in favour for limitarianism that do not require relational or altruistic reasonings from the side of those who have too much (Fourie 2022, 175; Huseby 2022, 239–241).

Next to such a non-relational reasoning, distinctively relational justifications for limitarian distributions can be advanced, too. This reasoning will be stronger, because moral concern for unequal distributions is triggered less by concern for those who have too much and more by concern for those who have less, often not even enough. Excessive accumulations of wealth among some constitute a morally indefensible waste of resources that could instead be used to satisfy unmet urgent needs of others (Robeyns 2017: 10–14). Failing to deploy resources to secure sufficiency and instead keeping them for one’s own, wasteful and potentially pointless pursuit of trivial improvements in well-being displays, in a harsh way, disrespect for others. It indicates that one’s own concern for one’s truly trivial pursuits weighs more than concern for others existential needs. No matter how widely this moral diagnosis applies, it challenges all those who find themselves above some such limitarian threshold. After all, it includes the significant moral reproach of being egotistic and enjoying one’s unjustified advantages even at the expense of other’s existential needs. To determine the size of the group of people subject to this criticism, much will depend upon where exactly to locate this threshold (cf. Robeyns et al. 2021).

A further relational argument for limiting the concentration of resources (possibly even under conditions where the basic needs of all are met) would highlight the detrimental consequences for egalitarian interactions and relations, in particular for political equality in a society. Anne Philips has shown the many forms in which (distributive) inequality, even if all have enough, still distorts (relational) equality, e.g. through separating spheres of daily interaction (such as housing, education, shopping etc.) according to different wealth-status, or through “delusions of superiority” that erode the ability to perceive and treat those who have less as equals (Phillips 2021, p. 82). Many other examples show how excessive concentrations of income and wealth distort political equality; e.g. when the affluent spend money to exploit mechanisms that transform money into political influence: through funding parties and campaigns, think tanks, media, and journalists, lobbyists etc. (cf. Christiano 2012). Even if, in principle, all can earn and spend money to this end, doing so comes cheaper the richer one is. A particularly extreme case is thus the influence of billionaires in world politics (Hägel 2020). All these arguments provide further support for Robeyn’s claim that limitarianism is instrumentally valuable, even decisive for achieving the intrinsically valuable goals of sufficiency and political equality (Robeyns 2017, p. 2).

Insofar as limitarianism challenges current injustices by focussing on those who would have to renounce advantages and in the end have less (instead of focussing on those who would deserve to receive more), it breaks with the widely accepted assumption in liberal capitalist societies that the accumulation of immense fortunes is acceptable if it is done by legal means and the result of, say, entrepreneurial skill and successful risk-taking. Limitarianism is thus greeted with particular resistance among those who are in this way privileged; and understandably so: after all, as a saying goes, When you are accustomed to privilege, equality feels like oppression. Yet, this does not undermine the strength of the arguments provided.

3.3 A Corridor of Legitimate Distributive Inequality. The Basic Formulation

Combining the two, partial demands of distributive justice – one sufficientarian and the other limitarian – now yields a first formulation of the distributive demands of relational equality in the form of a confined corridor of legitimate inequalities that provides necessary conditions for relational equality:

Distributions are just and supportive of relational equality if all have enough and no one has too much. Such a demand for a sufficientarian floor and a limitarian ceiling constitutes a “corridor” of legitimate inequality confined by two thresholds,

(a.1) all relating parties must have enough [sufficientarian floor]:

and

(a.2) no one must have too much [limitarian ceiling].

4 The Challenge of Residual Inequalities within the Corridor

The initial formulation of DTT presents highly ambitious demands, which are, given current circumstances, admittedly unrealistic and idealistic. Implementing these demands would eradicate the most extreme distributive inequalities. It would eliminate preventable material deprivation that cripples many lives and curtail excessive concentrations of wealth, both of which are symptomatic of and contribute to relational inequality. The demands can be applied to correcting distributive inequality both within societies, and globally (cf. Piketty 2021). Of course, such (re-) distribution of resources would not lead to (rough) equality in all possible dimensions: People would still be different in many significant ways, among other things regarding their well-being, physical abilities, luck in the pursuit of love or professional success. Yet, the relational aim of securing just distributions does not consist in equalising everything, only in reducing distributive inequalities to a degree compatible with and conducive to egalitarian relations.⁴ The initial attempt to limit inequality justifies imposing constraints on distributive inequalities that, while ambitious and unlikely to be implemented soon, remain moderate. These constraints align with the reasonable demands of pragmatic egalitarians who do not seek to equalize everything to the maximum extent possible. Some degree of inequality and difference will always be inevitable, or should not constitute an impediment to egalitarian relations.

But nevertheless: The corridor that might first result from a deliberation among equals willing to exercise egalitarian deliberative constraint, would allow for residual “intra-corridor” distributive inequalities that are still large, and possibly threatening for relational equality. As the following concrete examples will show, the two threshold concepts alone do not, yet, sufficiently reflect the reciprocal commitment to respect and treat one another as moral equals that lies at the heart of relational egalitarianism. Such a failure can be identified both on a group-specific structural or institutional level and on an interpersonal interactional level.

Example 1

CO₂-emissions may function as a proxy for material wealth and development/well-being (both globally and within countries). The distribution of anthropogenic greenhouse gas-emissions thus also reflects the extreme distributive inequalities that can be found within nations but also globally: The richest 10% of the world’s population are responsible for 50% of global emissions (with a disproportionate share resulting from the top 1%), while the bottom half of humanity generates less than 10% of global emissions (Chance and Piketty 2015; Oxfam 2020).

This, however, is not merely a fact about some random pattern of distributing the benefits (and harms) that come with CO₂-emissions, but in itself a symptom of unequal distributions in power and influence: The concentration of the heavy polluters among the former colonial powers in the global North and the concentration of low polluters among the former colonies in the global South clearly indicates a link between distributions and relational factors such as power and exploitation. Failing to take ambitious action by the privileged countries to self-limit and reduce harmful emissions, making space for and supporting increased emissions and development for others indicates a disrespectful attitude towards those who have less. And even if abject poverty were eradicated with generous support from the affluent countries so that all had enough in the countries of the global South, while citizens in the affluent countries of the global North would enjoy full flourishing (but without excess), this would not yet indicate genuinely egalitarian interactions. A merely voluntary, generous decision to support and redistribute (vs. acceptance of a duty in response to a legitimate entitlement) fails to acknowledge the equal moral status of the recipients of support (cf. Heilinger and Kempt, 2024, in press).

Example 2

The distribution of income and wealth can remain problematic, even if all had enough and no one had too much. Scheffler offers an illustration of the detrimental impact of unequal distributions upon interpersonal relations, even if neither partner suffers from genuine deprivation nor engages in wasteful life-styles. He writes: “If my wealthy friend regularly insists on going to more expensive restaurants than I can afford and on paying the bill for both of us, then I may feel that his allocative decisions […] are undermining the egalitarian character of our relationship.” (Scheffler 2015, p. 32) In this case, the choice of restaurant that, based on their unequal financial means, can be made by only one of the two friends morally taints the otherwise generous invitation. A significantly different range of available options thus constitutes in itself a challenge for egalitarian relationships. The rather trivial case of restaurant invitations is symptomatic for distorting effects on egalitarian relations that result from significantly diverging sets of available options enjoyed by one side, but unattainable for the other. Being aware of the fact that one may have significantly more or less than someone else is making relating on fully equal terms difficult, possibly even impossible.

Example 3

A two-tiered health-care system and unequal access to health care is another case in point (Fourie 2017; Nixon 2019). If some have significantly more (even if not unlimited) options to choose from, based on additional private insurance unavailable to others, this indicates and consolidates social hierarchies and differential status, generating not only unjustifiably different health outcomes but also proving detrimental to communities. Such challenges to relational equality persist, even if sufficiency is guaranteed for all, and if excessive treatments should be effectively prevented. Some individual leeway in choosing different health care plans might be acceptable from the perspective of relational egalitarianism, insofar as it may reflect individual preferences. But the dominating status quo in many countries where two equally needy persons receive differential treatment based on their different health insurance, is a clear indicator of distributive inequality that undermines, even violates social equality and perpetuates social stratification.⁵

The examples show that distributive inequalities, even if constrained by a sufficientarian and limitarian limit, may still fail to reach its aim of securing a stable basis for relational equality. Additional limitations on distributive inequality are necessary, especially regarding the distribution of existentially valuable goods and advantages, such as health care. Unequal control over resources or opportunities to achieve advantages undermines or complicates egalitarian interactions, as it invites or perpetuates subtle forms of domination and prevents adequate reciprocity. But what are these additional conditions that further specify the distributive preconditions of relational equality within the corridor?

5 The “Dual Threshold Theory” Refined

Given the shortcomings of the first attempt to define the distributive demands of relational equality with the help of a sufficientarian and a limitarian threshold concept, two further constraints to regulate intra-corridor inequalities should be considered to further reduce an unequal distributions’s negative impact on relational inequality: first, limiting distributive inequality through a particular ratio that further narrows the overly permissible first corridor; and second, requiring that one’s position within the resulting, available spectrum reflects one’s own responsible choices. Taken together, this set of demands spells out the distributive conditions that are not only necessary but sufficient to provide a sound distributive basis for relational equality.

As I will explain in the following, the first additional constraint further solidifies the basis of egalitarian interactions, the second secures the possibility of different personal preferences and life choices, while containing the impact of brute luck. These additional constraints are meant to apply generally in a lexical order: they have lower priority than the initial demand for implementing the wider corridor. (Starting from the current status quo, even securing minimal sufficiency for the worst off and imposing limits upon the best off is still an ambitious and remote goal that will, realistically speaking, not be reached any time soon.) But while the first set of demands should have overall more weight and priority, a context-sensitive and realistic application of the reasoning outlined in this paper should, of course, seek to make progress also on the second line of demands where possible, and not wait for full achievement of the first set of principles. Political reform and individual practice must start from the real conditions, which will nearly always be non-ideal.

Now, let us look at the two additional distributive demands of relational egalitarianism in turn.

5.1 A Context-Dependent Ratio of Legitimate Distributive Inequality

First, if the wide distributive corridor still negatively impacts upon relations, it should be narrowed down further. The adequate determination of the corridor’s width, however, will have to pay additional attention to distinctively relational considerations, resulting in a particular ratio of legitimate distributive inequality.⁶ The goal consists in identifying the right ratio of permissible inequality that does not undermine relational equality in a given context. Such a vague demand for a context-dependent ratio may not satisfy an ideal theorist’s desire for clear and unambiguous rules and principles, but it opens the theorising to for more context-dependent specifications and is thus crucial for making any theory relevant for practice. The actual work of finding the right ratio can only result from a knowledgeable and inclusive deliberation about the specific field under consideration which is shaped by the exercise of deliberative constraint. In the context of this paper, limit myself to some illustrative remarks about the examples introduced above.

Consider again the case of CO₂-emissions (example 1) that will not only generate benefits for their emitters but also negatively impact numerous others. Here it is difficult to justify, on the long run, any other standard than principally egalitarian rights to emit. Why should anyone have generally more rights to pollute and to impose costs upon others than anyone else? Accordingly, in recent political deliberations about emissions rights, the long term goal is nearly always to reach a (roughly) equal distribution. Some inequality, however, may be temporarily acceptable, as it is inevitable while transitioning to more egalitarian distributions. An egalitarian policy would thus have to implement a fair distribution of the remaining carbon budget that reflects the commitment to relational equality. Countries would have to undertake mitigation measures from their respective emission levels with the goal to eventually converge on (roughly) equal per capital emissions in all countries while realising overall declining emissions globally to reach net zero by mid-century (so-called “contract & converge”, cf. Stott 2012; Haines 2022). Only in taking this path will the distributive inequalities constitutive of relational inequality eventually be overcome.

Additional considerations may apply when determining the distribution of goods to people having special, costly needs to secure sufficiency and well-being: Those who depend upon additional emissions (e.g. for heating their homes in colder areas or for securing adequate levels mobility where this more costly than in other areas) should be granted permission. But regarding personal preferences, the distribution of emission rights must in principle be (roughly) egalitarian and remain altogether within the constraints imposed by the planetary boundaries that constitute the overall safe operating space for humanity (Rockström et al. 2009; Wilkinson/Pikett 2024). Permissible upper limits may thus be significantly higher or lower in particular contexts because of external constraints such as the climate, the unfolding climate crisis and planetary boundaries, than they would be on the initial relational justification of the corridor alone. Yet, also such indirect further narrowing of the corridor breathes a relational egalitarian spirit: Having and emitting in relation to what others have and emit is a matter of egalitarian respect.

Obviously, this type of reasoning holds true not only for the distribution of rights to impose costs upon others such as the environment, but also in terms of the distribution of positive resources to achieve well-being. If the ultimate value of relational equality consists in its assumed positive effect on the lives of people, this implies taking the interest of everyone to lead a good life – including using available technologies to this end – to weigh equally.

In the case of two friends dining out (example 2), the practical possibility to interact with one another and to reciprocate on (roughly) equal terms should be guaranteed, as well. This could be achieved e.g. in the form of limiting the acceptable ratio of inequality that may result from different incomes or wealth. One might, for example, stipulate that one person’s income and control over wealth must not exceed n-times that of the other person’s. Numerous strategies to achieve more distributive equality in societies are being discussed (and partly implemented) for centuries, among them progressive taxation (on income, wealth, or inheritances), unconditional income or a publicly provided inheritance for all (cf. e.g. Piketty 2020). Furthermore, specifically targeted strategies are being discussed, too, such as initiatives to reduce income inequality in corporations. Switzerland, for example, recently voted on – and rejected – a proposal to implement a 1:12 ratio between the highest and lowest paid employees in a company to reduce income inequality and secure fair wages (Graff 2022).

Interestingly, absolute equality seems not to be necessary, because some degree of income differentiation that may result from different functional roles in a hierarchic organisation with a division of labour does not necessarily undermine relational equality in an undue way. Keep in mind that functional hierarchies, as long as they do not generate class or status differences, should be tolerated also by relational egalitarians insofar as they allow through functional differentiation, sophistic cooperation and execution to better achieve the valuable goals of an institution. (It can also allow to financially reward different levels of effort and engagement, more about this in the following section.)

Furthermore, valuing a fancy dinner invitation does not only matter in monetary terms but also as an expression of generosity among friends or in the joint enjoyment of some activity. If this is what ultimately matters, reciprocating among pragmatic egalitarians does not presuppose exactly symmetric reciprocation. Inviting someone in return to a more modest restaurant or to any other enjoyable activity might function as perfect reciprocation and thus be fully in line with relational equality. And yet: Even if we admit for, or even welcome such multi-modal assessments, it is clear that the bigger the residual distributive inequality is, the more it will continue to constitute a challenge to egalitarian relations, demanding conscious navigation and management to uphold relational equality. The need for exploring and imposing context-specific ratios of acceptable distributive inequality thus remains.

In the case of health care (example 3), however, the corridor must be even narrower to reflect the higher moral relevance of health over recreational activities such as reciprocating dinner invitations. In existential matters, one even has to ask whether anything beyond equal distributions can be defended. A universal health care system providing everyone with the care that is necessary, available, and affordable when weighing the legitimate demands of all equally is ultimately the only acceptable solution fully aligned with the demands of relational equality. Even relatively small deviations that regard important health issues will advance not only a two-class health care system, but also a two-class society. Even countries with robust universal health care systems, such as Germany, which largely secure good care for all, are characterized by better and more expensive health care for those who can afford it and choose to opt out of the solidarity-based payment schemes of the universal health care system. If distributive inequalities in health care will be legitimate, they will have to be limited to non-essential services (say, a single-occupancy hospital room instead of one for two or three people) and remain significantly smaller than in other domains (such as income inequality with a permissible ration of, say, 1:12).

To sum this point up: From a relational perspective, the ratio of legitimate distributive inequality will primarily be determined by the impact it has upon the possibility to realise egalitarian interactions and relations. Here, of course, no one size fits all formulation of legitimate inequality can be offered. Importantly, must also differentiate between what would be ideally just and what would be a (remotely) realistic policy proposal for our current, markedly inegalitarian societies and world. Demanding exact equality in distributions, however, seems to be not required in many dimensions of human lives and living together as long as the residual inequality does not exceed a specific ratio.

5.2 Equal Opportunity and Responsible Choices

Now, would such a ratio alone fully satisfy all justifiable distributive demands a relational egalitarian may raise? Realistically speaking, much would already be gained if the first formulation of the corridor were ever to be implemented; much more would be gained if residual inequalities were confined by such context-specific ratios to be determined in joint deliberation. But all these demands are “only” external demands that apply unconditionally and limit inequalities while somehow failing to honour the autonomous and potentially quite different preferences and choices people may have and make. If there is a ratio of distributive inequality that allows for different positions on a spectrum, and if the position one finds oneself in can be determined by individual preferences and choice, than one’s position should ideally and wherever possible depend on that choice and not on factors beyond one’s control. A certain element of luck-egalitarian thinking thus can be integrated into specifying the distributive demands of relational egalitarianism – and it should. Honouring everyone’s status as a moral equal requires also respect for individual choices and preferences. Should someone value to pursue certain goods and thus have more, should someone be willing to work harder⁷ or to take calculated gambles to have more even if it might come with a risk to have less, if some would behave more or less responsible and if all would have (roughly) equal opportunities to behave in one way or the other, then why not allow for some degree of differential outcomes – as long as these are certain not to negatively impact neither essential dimensions of personal lives nor the ability to fully relate to others as moral equals? In my view, the burden of proof here shifts towards those who would demand even more distributive equality than the one that will be realised by the conditions mentioned so far, amended by this second intra-corridor condition requiring that differential outcomes result from responsible choices.

To sum up this point: Residual intra-corridor inequalities do not conflict with relational equality if they are kept within a particular, context-dependent ratio, and if they result from people’s responsible choices. Keeping some degree of difference, also in distributive terms, is also valuable from a relational perspective: Relational egalitarians respect, even welcome difference, individual preferences and the choices people may make in this light: As long as distributive inequalities result from such choices – and do not exceed a particular ratio –, they do not undermine relational equality.⁸ For this to be true, however, it must be secured that the set of options initially available to people is roughly equal, and that factors that are beyond control of individuals (and cannot be equalised by social reform such as distributions in the natural lottery) do not translate into disadvantage or inferior status. Once these demands are met, pragmatic egalitarians will be able to live with the resulting distributive inequalities, and focus their attention on practicing relational equality well.

5.3 The Corridor of Legitimate Distributive Inequality

Thus, we can now summarise the distributive demands of relational egalitarianism by expanding the initial formulation of the – still too wide – corridor of distributive inequalities. The full formula of the Dual Threshold Theory (DTT) thus demands that, once the lexically prior sufficientarian and limitarian principles (a.1 and a.2) are met, two additional conditions apply:

Residual intra-corridor distributive inequalities must

(b.1) remain within a particular ratio that reflects in a context-dependent way the moral relevance of the respective distribution [contextual demand];

and

(b.2) result from responsible choices of individuals [equality of opportunity demand].

If these demands – a.1 and a.2 as well as b.1 and b.2 – are met, a sound distributive basis for relational equality exists; or so I have argued.

6 Concluding Remarks

I have determined the distributive demands of relational equality by proposing a dynamic corridor of legitimate distributive inequality, confined first by a sufficientarian and a limitarian outer threshold, within which, second, a particular, context-dependent ratio of acceptable inequality may persist under conditions of equality of opportunity. The justification for this determination lies in the exercise of egalitarian deliberative constraint among those who understand themselves as moral equals and seek to realise relations and interactions on a footing of equality.

To conclude, I want to briefly mention five objections against the proposed account and at least hint towards possible refutations.

A critique may doubt (objection 1) the assumed impossibility of egalitarian relationships if people have less than enough. Anecdotical evidence from disaster settings, where no one has enough, shows the possibility of thriving solidarity and relational equality among the deprived. And indeed, relational equality can possibly reign among the deprived as it can among the (super-) privileged. But this observation rather supports than contradicts the current proposal: It is a degree of relative distributive equality that matters as a basis for egalitarian relations, and since it is neither desirable nor necessary in a world of abundance to limit all to a state of deprivation, nor is it realistically possible to elevate everyone to a level of supra-limitarian affluence, reasonable efforts to provide a distributive basis for relational equality should first target the identified corridor and then limit the residual intra-corridor inequalities as outlined above.

Critiques have challenged (objection 2) threshold concepts for being somewhat arbitrary, and indeed it is not trivial to identify what exactly counts as “enough” or “too much.” Efforts have been made to provide arguments for context-dependent specifications of both thresholds (e.g. Axelsen and Nielsen 2015; Robeyns et al. 2021), and this remains an ongoing task, given the ever-changing state of socio-economic affairs in different places and times. Related is the question whether it makes sense in the first place to have a discussion about limiting distributive inequalities without pre-selecting and sticking to a single, distinctive metric of justice, e.g. resources, or well-being, or capabilities/opportunities? After all, different metrics generate distinctive problems sui generis, which may make general talk about distributions insufficiently specific. I read this important challenge as an invitation for further research to test and advance the proposal by making it more specific and relevant in concrete contexts.

Another objection might label my proposal as mere “ideal theory” with little or no probability to be widely implemented any time soon (objection 3). And while I share the doubt about its future realisation, the proposal itself is not built on ambitious or extravagant foundations, only on the assumption that all humans do matter equally/have equal moral status and deserve equal moral standing. If this rather modest normative investment generates distributive demands that are perceived as excessive, one is left with two options (besides refuting the proposal outright): Either one bites the bullet and accepts the apparently excessive demands as justified, or one reduces the firmness of one’s commitment to the modest normative investment.

I chose the former option. Indeed, many of the distributive schemes from which not only the super-affluent but also many “middle class” people benefit cannot be justified if one takes the egalitarian deliberative constraint minimally seriously. Rejecting my proposal for being excessively demanding towards those who benefit from inequality then appears as a self-serving, hypocrite strategy to deflect responsibility and to suppress feelings of responsibility and guilt. It needs to be acknowledged that the many benefits of structural (relational) injustices come with massive distributive inequalities, even if this knowledge is uncomfortable for the privileged and generates very demanding moral obligations.

A critic may also wonder (objection 4), why my proposal did not start with and limit itself to demanding genuinely fair equality of opportunity (demand b.2 above), as this ultimately constitutes the core of the proposal which in itself may be sufficient to satisfy the distributive demands of relational equality. Such a lean, luck-egalitarian demand alone would indeed capture a minimal condition for relational equality. In particular it would be immune to objection 1 as it acknowledges that equality can persist in the case of communities shaped by deprivation or abundance. Yet, these are special cases, and the general goals of relational egalitarianism, as outlined above, importantly include ending deprivation as well as extreme abundance, as they are considered both intrinsically and instrumentally wrong and detrimental to egalitarian relations. A demand of fair equality of opportunity alone would fail to acknowledge these important concerns.

A related but quite different challenge (objection 5) may question the need to include fair equality of opportunity in the proposal: Is it really necessary for relational equality that one’s location on the narrow distributive spectrum permitted by the refined corridor results from one’s own responsible choices?⁹ Some may find themselves within the corridor as a result of parental gifts, others because of carefully saving every penny, yet others because of transfer payments provided by the community. Why should such different causal dynamics undermine egalitarian relationships? And indeed, the sheer knowledge that all will end up roughly equally well off in distributive terms no matter their social background or personal behaviour might already be sufficient for not undermining relational equality. But given that the proposed legitimate ratio of distributive inequality is to remain non-trivial (say 1:12 to accommodate space for different preferences, choices, and outcomes) and given that group-based distributive inequalities based on gender, ethnicity, social class etc. have to be eliminated for relational equality, equality of opportunity remains essential for securing a sound basis for a community of equals.

Obviously, more work is needed to fully assess the flaws and merits of the proposed account, in particular when it comes to its practical use. But this task is beyond the scope of the present paper which only set out to provide the contours of a first full outline of the distributive demands of relational egalitarianism.¹⁰