2. Limitarianism

In her groundbreaking paper, Ingrid Robeyns (Reference Robeyns, Knight and Schwartzberg2017) offers a defense of limitarianism—the view that given certain conditions, it is morally impermissible to “have too much.” While the position according to which excessive wealth is a moral vice is not a new one (Kramm & Robeyns, Reference Kramm and Robeyns2020), Robeyns’ paper triggered a debate among political philosophers about whether extreme wealth also gives rise to duties of justice, and what those duties consist in (Volacu & Dumitru, Reference Volacu and Dumitru2019; Zwarthoed, Reference Zwarthoed2018; Timmer, Reference Timmer2019, Reference Timmer2021a, Reference Timmer2021b; Ali & Caranti, Reference Ali and Caranti2021; Ali, Reference Ali2022; Huseby, Reference Huseby2022; Hickey, Reference Hickey and Robeyns2023b; Meijers, Reference Meijers and Robeyns2023; Icardi, Reference Icardi and Robeyns2023; Neuhäuser, Reference Neuhäuser and Robeyns2023; Berkey, Reference Berkey2022; Nielsen & Axelsen, Reference Nielsen and Axelsen2022; Halldenius, Reference Halldenius2022; Flanigan & Freiman, Reference Flanigan and Freiman2022). Theories of distributive justice typically focus on those who have too little and have largely overlooked questions concerning the top end of the distribution. This is the unique contribution of limitarianism—suggesting that above a certain point, unique moral duties apply, and justice requires taxing all excessive wealth (Robeyns, Reference Robeyns, Knight and Schwartzberg2017).

Robeyns puts forward two justifications for limitarianism, conceding that there may be further justifications: First, Robeyns argues, great inequalities in wealth undermine political equality (Robeyns, Reference Robeyns, Knight and Schwartzberg2017). The ultra-rich use their resources in a variety of ways to acquire political power and affect political decision-making. They fund political parties and candidates, and affect public opinion by buying media outlets and hiring lobbyists. They can even direct the creation of knowledge to serve their political ends by funding research that provides scientific credence to their opinions. Additionally, having enough resources sometimes enables the ultra-rich to evade policies they are unable to block. No institutional rule, Robeyns argues, is capable of isolating the spheres so that money will not spillover to politics, therefore people have a duty of justice to other people not to have too much. Moreover, extreme wealth may create political inequality even when resources are not used to affect politics since wealth can feed into relations of domination (Timmer, Reference Timmer2019).

The second justification Robeyns offers for limitarianism is that it is immoral for people to be rich when other people have unmet urgent needs (Robeyns, Reference Robeyns, Knight and Schwartzberg2017). Limitarianism depends, therefore, on a set of empirical conditions concerning the existence of extreme poverty, extreme disadvantage, or collective action problems, that can be treated by governments using financial resources. One specific context in which such needs currently present themselves is the global ecological crisis, the alleviation of which is frustrated because of collective action problems and lack of resources (Robeyns, Reference Robeyns2019; Buch-Hansen & Koch, Reference Buch-Hansen and Koch2019; Green, Reference Green2021; Hickey, Reference Hickey2023a).

As Robeyns concedes, there may be additional justifications for limitarianism. One such justification is offered by Zwarthoed (Reference Zwarthoed2018), who argues that excessive wealth is “detrimental to the development and exercise of rich people’s autonomy.” Neuhäuser (Reference Neuhäuser and Robeyns2023) offers an additional justification, according to which having self-respect relies on securing equal standing, which is impossible in a society with severe economic inequality.

Given the plurality of justifications for limitarianism, there may be cases in which only some of the justifications apply, and in which a threshold is compatible with one and not the other (Ali & Caranti, Reference Ali and Caranti2021). In later writing Robeyns acknowledges this (Reference Robeyns2022), stating that in such cases limitarians will have to “work out how to deal with multiple limitarian thresholds” (260).

There is some debate concerning the question which kind of principle of justice the principle of limitarianism is. Understood as a comprehensive theory of justice it entails that satisfying the requirements of limitarianism is all that is required in terms of justice. In other words, assuming the preconditions obtain, once wealth above the threshold (however defined) is taxed, justice obtains. The contemporary work, however, accords limitarianism a more limited role within distributive justice, either as one component of a pluralist theory of justice (Dumitru, Reference Dumitru2020; Zwarthoed, Reference Zwarthoed2018; Robeyns, Reference Robeyns2022; Halldenius, Reference Halldenius2022), as a mid-level principle, that can inform policy making (Robeyns, Reference Robeyns, Knight and Schwartzberg2017; Green, Reference Green2021; Timmer, Reference Timmer2021b; Neuhäuser, Reference Neuhäuser and Robeyns2023), or as a presumption (Timmer, Reference Timmer2021b; Huseby Reference Huseby2022). Thus, wealth-limiting policies could be endorsed, for different reasons, by proponents of different comprehensive theories of justice. An attractive account of limitarianism should, therefore, be compatible with other distributive principles.

The nature of the threshold is a crucial factor in a limitarian principle of justice. The threshold characterized by Robeyns involves an “absolute threshold” drawn at the point of satiation of human flourishing. To put it differently, the metric of justice suggested is human flourishing, and the principle of distribution that defines the threshold is full realization or satiation: once a person has achieved a fully flourishing life, surplus resources are morally inconsequential.

Robeyns imports the notion of flourishing from the capability approach, however for the sake of limitarianism, Robeyns holds that the threshold is meant to capture the material side of quality of life and that it excludes valuable non-material goods such as “one’s opportunities to be active in local politics, or the capabilities of being part of a religious community and engaging in its practices.” (Robeyns, Reference Robeyns, Knight and Schwartzberg2017, 26). A fully flourishing life is not a matter of subjective desires or tastes; nor does it depend on one’s relative wealth compared to others; it requires fulfilling an objective list of elements, such as nourishment, access to health care, education, transportation, and housing. Although the components of a flourishing life are an objective list that applies to all humans, the list is also context-dependent in that the details of what it means to fulfill it are determined with reference to time and place. Take, for example, the human functioning of mobility: in order to live a fully flourishing life in 21st century Europe, people require: “access to the goods that enable one to be mobile within a radar of a few hundred miles” either by car or public transportation. But, Robeyns argues, “one wouldn’t need to have access to a private jet” (Robeyns, Reference Robeyns, Knight and Schwartzberg2017, 26).

Although Robeyns’ threshold is conceptualized in terms of human flourishing, resources are used as a proxy for human flourishing. The various components of human flourishing are detailed and examined, to determine what is required to satiate each of them in the specific social and economic context and calculate the cost of obtaining them. The result is an absolute sum above which an individual in this society is too rich. Any resources that exceed the overall sum are surplus resources.

Importantly, it is perfectly possible to lead a fully flourishing life, having satiated the objective list of elements needed for human flourishing and consequently reached the riches threshold, despite having unmet (subjective) desires. Even the ultra-rich have desires that require resources to fill, though they are not what we would call “urgent,” such as the desire to buy a new private jet or mansion. Satisfying these desires, according to Robeyns, does not increase flourishing—which is logically impossible once people are fully flourishing. The inverse is also true: people may feel content without leading fully flourishing lives according to Robeyns’ objective list. And since what matters to Robeyns in terms of justice is human flourishing, surplus resources that cannot contribute to flourishing have “zero moral weight” and taking them is morally unproblematic.

Applying this straightforwardly would require 100% taxation on all surplus resources. However, Robeyns is willing to relax this requirement and design a tax policy so that rich people maintain incentives to create resources needed to alleviate poverty. Robeyns makes even more far-reaching adjustments to the resources we can tax in a later restatement of limitarianism (Reference Robeyns2022). Robeyns recognizes that “the limitarian threshold, above which we take away and reallocate money, and the riches line, being the line at which surplus wealth kicks in, need not be the same” (254). This will be discussed later on.

Other means of determining thresholds have also been suggested. In seeking a threshold suitable for autonomy-based limitarianism, Zwarthoed (Reference Zwarthoed2018) examines the mechanisms that give rise to autonomy deficits among the wealthy and seeks to limit them, suggesting we rely on popular opinion for setting the threshold. Empirical work examining popular attitudes toward excessive wealth shows that many people do indeed think that excessive wealth is immoral and are even able to specify the monetary threshold (Robeyns et al., Reference Robeyns2021).

I now argue that Robeyns’ account of the limitarian threshold is unattractive and that a relative threshold account is more persuasive. A relative threshold, for the purpose of this paper, means that the resources one has to have to be considered “too rich” are determined in relation to the wealth of others in society, most notably to those who have the least, but possibly also to others (for example, excessive wealth might be determined in relation to average wealth). I argue that relative thresholds are better aligned with the two traditional justifications of the theory (part 2); that they avoid the powerful incentives objection (3); that they provide an account that does not rely on robust normative assumptions and are therefore more suitable for a partial theory of justice (4), and finally that relative thresholds sit well with limitarianism’s underlying intuition (5). Initial comments concerning the specific design of relative thresholds will be made in the conclusion.

3. Relative thresholds better align the justifications of limitarianism

I argue that an attractive account of limitarianism is one in which the threshold is compatible with the justification, namely likely to promote the goals that justify limiting wealth. However, it does not necessarily need to fit the justifications precisely. In fact, it would be quite difficult to design a threshold that would continuously fit the justification, since redistribution, voluntary exchanges, and other social contingencies constantly affect the level at which the threshold would fully realize its justification. For example, attending to the precise scope of unmet urgent needs entails continuously adjusting the threshold according to fluctuating housing prices, inflation, and more. Additionally, taxes imposed due to other principles of justice may apply alongside limitarian taxes, thereby affecting relative and absolute levels of wealth (for both the rich and the poor) in unpredictable ways. As a result, attempts to provide a threshold that is both precise and stable are challenging, and full compatibility with the principle’s justification is not necessary.

Still, to do their work, namely ground the requirement for special taxation of the rich, justifications must have some relationship to the threshold that defines “the rich.” For limitarianism to be operable, even as a policy guiding principle, thresholds need justification. Therefore, whatever the role we assign limitarianism within theories of justice, an account that aligns the threshold with the theory’s justifications is preferable.

Furthermore, several different justifications have been offered in support of limiting wealth, and others may be available. It does not seem likely, nor necessary, that a single threshold will be compatible with all the justifications. Therefore, I argue that limitarianism is best seen as a set of restrictions on wealth—each with its own justification, threshold, and conditions for application. And one can endorse some or all limitarian principles, with the ensuing normative implications. While Robeyns hints that there may be different justifications that lead to different thresholds, she stops short of stating this explicitly and assumes instead that these are differences within the theory that should be resolved.

The discussion here focuses on the two traditional justifications for limitarianism, arguing that a relative threshold is better suited for both. Other limitarian principles, based on their different justifications, might well require different kinds of thresholds, including thresholds that are not determined relatively.

Robeyns’ flourishing satiation threshold undoubtedly contributes to promoting the two traditional justifications, therefore the necessary relation between threshold and justification could, theoretically, be satisfied. However, as I now argue, a relative threshold is much better aligned with both justifications.

Beginning with the first justification, namely preventing political inequality: to ensure that excessive wealth does not jeopardize political equality, one could argue, that the threshold should be set at the point where this risk materializes. However, there is no reason to assume that this will always (or ever) be the point at which the rich are fully flourishing (see also Halldenius, Reference Halldenius2022). Significant political inequality could happen even when people do not have enough resources to lead a fully flourishing life; and conversely, political inequality, of the kind and severity that limitarians aim to limit, might happen only in cases of extreme wealth, that far exceeds satiation. Funding a think tank or supporting scientific research about climate change, for example, are actions that are only available to a handful of exceptionally wealthy people, whereas many people who lead fully flourishing lives are probably unable to fund such endeavors.

Of course, people who are well off but not extraordinarily rich can also organize and promote their political goals. Whether this is unjust or not would depend on whether individuals amass more than their fair share of power or whether such organizing is possible for a wide range of individuals and groups. The ability to do so by one’s self is unique to the ultra-rich.

The flourishing-satiation threshold is also not directly linked to realizing unmet urgent needs, which is Robeyns’ second justification for limitarianism (see also Halldenius, Reference Halldenius2022). In a relatively well-off society, taxing 100% above the threshold might result in raising more revenue in taxes than needed for meeting all urgent needs, casting doubt on whether there are moral grounds for levying such a drastic tax (although, of course, taxation might be justified to prevent political inequality, or for other, non-limitarian reasons). The flourishing-satiation threshold could also result in insufficient revenue for realizing unmet urgent needs, although this is less problematic for Robeyns considering that limitarianism is but one component of justice, and additional principles can justify taxation of resources below the satiation threshold.

In response to this critique, in a recent restatement of limitarianism (Reference Robeyns2022) Robeyns concedes that there may be a gap between the riches line, namely the point at which flourishing is satiated, and the limitarian threshold, which is the point above which resources should be taxed. The latter should be determined in accordance with the relevant justification for limitarianism.

Separating the riches line and the limitarian threshold seems like a helpful move in response to my criticism. However, it is cumbersome, adding another threshold to the already quite complicated limitarian account that Robeyns puts forward. Also, this move raises a new set of pressing questions that Robeyns does not address: what are the relations between the riches line and the limitarian threshold? is crossing the riches line a precondition for limitarian taxation? why is such a precondition necessary? Are rich people allowed to keep surplus resources if the justified threshold is higher than the riches line? and, importantly, does this mean that limitarianism is actually a set of independent limitarian principles?

But Robeyns’ solution falls short even more fundamentally because the justifications for limitarianism should not merely tell us when excessive wealth should be taxed; instead, the justifications are what render wealth above the threshold immoral. Having more resources that are needed for living a fully flourishing life is not, in itself, immoral, even according to Robeyns. Rather, immorality depends on other facts, such as the existence of unmet urgent needs or political inequality.Footnote ³ An attractive version of limitarianism, therefore, is one in which the definition of surplus resources aligns with the justification for taxing them.

A relative threshold account of limitarianism does not run into these problems. The political inequality justification is inherently comparative and is therefore compatible with a relative threshold. The dangers to political equality Robeyns describes as the basis for limitarianism are caused by large material inequalities, irrespective of objective holdings (Ali, Reference Ali2022; Huseby, Reference Huseby2022). Like other positional goods whose value is determined comparatively (Brighouse & Swift, Reference Brighouse and Swift2006), influencing politics requires not only having resources but also having more resources than one’s opponent. Money can buy media outlets, promote candidates and affect the direction of research, but as long as it is offset by similar investment from one’s political opponents (or at least the possibility of others to offset these influences, if they so choose), political equality of the relevant kind is maintained. The role that wealth plays in enabling political inequality, to sum up, is not a function of excessive wealth as such, but of large material inequality.Footnote ⁴

Furthermore, large inequalities, of the kind that hinder political equality, and therefore should trigger limitarian concern, may exist even when people have not reached the flourishing-satiation threshold. Assume, for example, that people reach satiation (ST = Satiation Threshold) at 100 and that inequality above which political equality is impossible is when the wealthy have more than 25 times the resources that the poor have. In case A (5,125), the rich are above the satiation threshold, and would therefore be taxed on both accounts of limitarianism—the objective satiation threshold (100) and the relative (1:25) threshold. Now think of society B (3, 75). The richest member is below the satiation threshold, which is 100, but the kind of inequality that jeopardizes political equality exists. In both cases material inequality is large enough to undermine political equality, suggesting that limitarianism should tax both, yet Robeyns’ limitarian threshold only applies to case A.Footnote ⁵ Determining the threshold relatively would capture both, and inequalities that prevent societies from obtaining political equality would be prevented in both cases. This example demonstrates that a relative threshold better addresses political inequality. To further complicate things, the threshold represents the sum of the resources needed for full flourishing in all of life’s spheres. People, however, can decide to use their resources for different things, for example, using some of the resources designated in the model for transportation for other ends such as housing. As a result, people might be able to obtain political dominance without exceeding the riches threshold.

We could, of course, argue that while all cases of political inequality raise valid concerns of justice, only when these occur above the satiation of flourishing they are concerns of limitarian justice. All other cases may be addressed through other principles of justice. While true, this detracts from the attractiveness of limitarianism since it does not account for the different responses in similar cases, and also makes limitarianism less important from a practical standpoint.

I argue that the relative threshold is better attuned to the second justification too. As stated above, taxing all resources above full flourishing does not necessarily provide the right amount of resources to address unmet urgent needs.

Rejecting Robeyns’ account of the threshold does not yet show that the alternative account that I propose, namely the relative threshold is compatible with the urgent needs justification. If limitarianism is meant to supply the resources for satisfying urgent needs, the straightforward way to define the threshold is to perform an audit of the needs and resources in society and just take what is needed. A relative threshold does not necessarily fare better than a flourishing satiation threshold in this regard. But we should take into consideration that an accurate evaluation of needs and resources is ephemeral and would stop being relevant as soon as any exchange or redistribution occurs. There are, therefore, advantages to setting less accurate thresholds that are applicable for longer periods of time. Moreover, there may be additional, non-limitarian, requirements of justice that can contribute to the resources available for redistribution. With these caveats in mind, I argue that a relative threshold is better suited for capturing the notion of excessive wealth.

Robeyns characterizes flourishing as contextual, meaning that determining what is needed for fulfilling needs (and satiating human flourishing) requires knowing what goods and services are available in a specific society at a specific time, and how much they cost. On the other hand, we seldom think about urgent needs as something that is determined relatively. But the very definition of urgent needs, and not only their cost, at least in the relevant sense for limitarianism, I argue, is largely contextual (see also Nielsen & Axelsen, Reference Nielsen and Axelsen2022).

Recall, that the reality of unmet urgent needs is treated as a justification for the drastic measure of fully taxing surplus resources. As such, urgent needs are not determined only by examining facts—which needs are crucial for human life, but also normatively: which needs, when unmet, would justify limitarian taxation. Therefore, there may be needs that are crucial for human survival, but nonetheless, their absence does not constitute an injustice. For example, although food and water are crucial human needs, if a famine strikes and food and water are not available to anyone at all, we would not say that starving is unjust. On the other hand, needs may be urgent and fulfilling them is needed for justice even though they are not crucial for bare survival.

Few would dispute, to give another example, that access to electricity, vaccinations or sanitation is crucial for human flourishing. However, for most of history electricity, vaccinations, and modern sanitation did not exist, and even after they were invented, they were only available to a privileged minority. In those times, despite being detrimental to chances of survival, not having these goods did not constitute an injustice. Only when these things became easily available to almost everyone (because they became inexpensive, and there is no objective shortage of them) did their absence elicit the kind of moral claim associated with “urgent needs” in limitarian theory. Like Adam Smith (Reference Smith1776) teaches us, since many goods have positional dimensions, what we consider a mark of extreme deprivation depends on what others have, not merely its importance for human life.

“A linen shirt […] is, strictly speaking, not a necessity of life […] But in the present times, through the greater part of Europe, a creditable day-labourer would be ashamed to appear in public without a linen shirt, the want of which would be supposed to denote that disgraceful degree of poverty…”

Many material goods have positional aspects, meaning that their bearing on a person’s relative standing can confer dignity or disgrace. Therefore, there may be things that have profound effects on human flourishing, indeed are needed for their very survival, that are not “urgent needs” in the morally relevant sense (namely, they do not create the strong moral claim that gives rise to limitarian justice). And, like the linen shirt, things that are not, strictly speaking, needed for sustaining life, but are socially accepted, can be an “urgent need,” and meeting them justifies taxing the rich at 100%.

The definition of “urgent needs” that follows from this analysis is comprised of two components: first that they have a crucial effect on human flourishing, and second, that they are very widely accessible, or at least could easily be made available to all. So if all the resources in society are being hoarded by one extremely wealthy man, and everyone else is starving, we would say that food is an “urgent need” despite not being widely accessible, because his wealth could feed everyone else. The determination of urgent needs according to these two components is non-binary, in other words, the more people have access to a certain good, and the more important it is for human wellbeing, the worse the injustice in its absence. In any case, defining urgent needs, I argue, requires looking at what other people have.

A life-saving drug, for example, or an atomic shelter, may be crucial for human survival. However, if only the ultra-rich can afford them, and they are only theoretically available for most people, they would not be considered an urgent need for the sake of limitarianism, and consequently, people who do not get them do not suffer injustice, merely an unfortune. Conversely, urgent needs are what people take for granted and feel that without them they cannot lead a decent life. Thus, if average longevity were 45 years, those who lived to the age of 48 would typically not dream of complaining that they have unmet urgent needs in areas of nutrition, health care, or sanitation. On the other hand, when people around the world live well into their eighties, dying at 48 due to lack of adequate health care might reasonably ground a claim of severe deprivation.

Needs such as telephones, computers, and internet connection are an interesting case for how we think about urgent needs. Clearly, owning the most up-to-date smartphone does not qualify as an urgent human need, but functioning telecommunication infrastructure and services such as phone and internet connection, as well as devices such as computers, are becoming essential for participating in society and are therefore increasingly viewed as urgent needs.

I should stress that while urgent human needs are contextual, the content of the list of urgent needs is objective in the sense that it does not depend on any individual’s subjective preferences. Needs are urgent even if specific individuals choose not to fulfill them, although if many people decide to relinquish certain goods, this would probably indicate that they do not, in fact, constitute urgent needs.

A relative threshold, admittedly, is not logically derived from this justification. Urgent needs were defined as those that are important for human flourishing and that are, or could easily be, almost universally accessible, whereas the proposed threshold is derived from the relation between the resources available to the rich and those available to the poor. The threshold will, most probably, not correspond exactly with the sum needed to satisfy urgent needs. Still, a relative threshold better expresses the contextual nature of urgent needs than a flourishing threshold and is flexible enough to incorporate, as society becomes richer, more services and goods within the scope of “urgent needs.”

From a more practical point of view, a relative threshold ties the material position of the rich to that of the poor. If the rich wish to accumulate resources, they will be incentivized to improve the situation of the worst-off or else be taxed. The 2016 Israeli law pertaining to the compensation of senior managers demonstrates the potential influence that a ratio threshold can have: its most notable result, perhaps unexpectedly, was not to decrease the salaries of senior managers. Rather, it raised contract workers’ wages, thereby enabling the seniors’ wages to remain high. A relative riches threshold could potentially have the same effect of improving the condition of the poor, allowing them to enjoy the prosperity of the wealthy.

4. Relative thresholds maintain incentives

The initial proposal put forward by Robeyns involves taxing 100% of the resources above the riches threshold, since these have, according to Robeyns, “zero moral weight.” Imposing an absolute tax, however, would disincentivize the wealthy to be productive above the threshold, which is inefficient, and more worryingly for limitarianism—it is counterproductive for justice because the surplus resources that were supposed to realize the poor’s urgent needs might not be created (Volacu & Dumitru, Reference Volacu and Dumitru2019).

Responding to this objection, known as the ‘incentives objection’, Volacu and Dumitru (Reference Volacu and Dumitru2019) argue for a weaker version of limitarianism which caters to responsibility and calls for revenue-maximizing tax policy. Robeyns (Reference Robeyns, Knight and Schwartzberg2017) also recognizes the strength of this challenge and concedes that while limitarianism ideally justifies 100% taxation, the actual tax will be set at a lower level that will preserve rich people’s incentives. Timmer (Reference Timmer2019) suggests thinking beyond income tax and designing limitarian policies that do not run into this problem in the first place, for instance imposing an inheritance tax that allows people to bequeath their wealth only if it exceeds what they inherited from their parents.

While incentive-sensitive limitarianism offers a practical solution for the challenge, I argue that a solution that aligns what we define as “surplus resources” with what we can fully tax is more attractive due to its simplicity and coherence: the threshold sets the moral limits, and all resources above it can be taxed at 100%.

Under the relative threshold, the rich are, in principle, allowed to accumulate unlimited wealth as long as the less advantaged members of society “keep up” and become better off, allowing no more than the permissible gap between the richest and the poor to obtain (Ali & Caranti, Reference Ali and Caranti2021; Ali, Reference Ali2022). The rich would, therefore, be incentivized to give out higher paychecks to the people they employ and to otherwise invest in promoting society’s weakest links. Additionally, the redistribution of surplus resources that would occur as a result of activating the limitarian threshold, would itself, raise the riches threshold, by improving the situation of the worst-off. As a result, a relative threshold is likely to improve the objective financial condition of the worst-off, while incentivizing the wealthy to keep producing.Footnote ⁶

5. Relative thresholds are suitable for a partial principle of justice

Another advantage of a limitarian account that employs relative thresholds is that it does not require committing to a robust set of normative and empirical assumptions, and is, therefore, better suited to work in concert with other principles of justice. Since it is viewed by most writers as a partial principle of justice (but see Halldenius, Reference Halldenius2022), an attractive version of limitarianism is one that can be applied together with different theories of justice and is compatible with different theories of the good life.

The traditional view of limitarianism can apply alongside principles of justice that are directed below the threshold, as Robeyns (Reference Robeyns2022) spells out in detail; however, the threshold itself is defined in terms of flourishing which is a contested concept, so many might not be able to endorse limitarianism. This, I argue, is regrettable, and a preferable threshold would be as normatively ‘thin’ as possible.

In what follows I point out three aspects of the flourishing satiation threshold that require endorsing contested normative assumptions: the concept of human flourishing, the possibility of satiate flourishing, and the moral value of surplus resources.

The first aspect involves the concept of human flourishing. The limitarian threshold, according to Robeyns, is set at the point at which human flourishing is satiated, determined by an objective list version of human well-being imported from the capability approach, which she defends extensively elsewhere (Robeyns, Reference Robeyns2017a). The capability approach and objective approaches to well-being more generally have their merits, no doubt, but there are clearly other plausible approaches to well-being, and several different variations of objective list theories (Arneson, Reference Arneson1999; Rice, Reference Rice2013; Fletcher, Reference Fletcher2013). Ali & Caranti (Reference Ali and Caranti2021), for example, put forward several possible objections to the way Robeyns constructs the threshold using the capabilities approach (see also Ali, Reference Ali2022). I find some of these objections convincing, but my argument lies elsewhere, namely in the fact that endorsing Robeyns’ limitarian theory requires taking on board many normative and methodological commitments. This, clearly, does not refute Robeyns’ account, but it does make it difficult to see how a variety of different theories (of value and justice) could work with it, which is one of the theory’s desiderata. A resource-based relative threshold avoids the challenge related to the thick concept of flourishing that is assumed by Robeyns’ account of limitarianism. It does not require us to agree on what things contribute to human flourishing, and on the relations between flourishing and resources. It merely requires acknowledging that resources contribute to various life projects and conceptions of the good and are therefore an appropriate subject for limitarian justice. And while objections may be made to this modest premise too, it is much less contested than those put forward by Robeyns.

Another issue around which disagreement may arise in the traditional limitarian account involves the satiation of human flourishing. The riches line is set at the point at which people have enough resources to live fully flourishing lives. Further resources, according to Robeyns, have zero moral weight. And while Robeyns’ threshold is determined in material resources it relies on the possibility that human flourishing is satiable. However, there can be significant disagreement regarding the statement that human flourishing in general, and each of the specific functions, can be satiated, and if so, at what level satiation occurs.Footnote ⁷ It is unclear, for example, what satiation would mean in the context of education. Most likely, this would consist of a sufficientarian account or an opportunity account of education. While both are legitimate accounts of educational justice for the lower threshold, there are also reasonable theories of educational justice that would argue, contra Robeynsian limitarianism, that investing more in education above the threshold has moral value, even if there are unmet urgent needs and if it worsens political inequality.Footnote ⁸ Education is but one example; similar disagreements could involve other functionings too, resulting in people rejecting limitarianism for the “wrong” reasons. A relative threshold account of limitarianism does not need to take on any commitments concerning human flourishing, and as noted, it merely requires conceding that on any reasonable account of flourishing resources have instrumental value.

Finally, traditional limitarianism assumes that surplus resources have “zero moral weight” and therefore taking them is morally unproblematic. The riches line represents the resources that are needed to lead a fully flourishing life in a specific society, and at which human flourishing cannot be enhanced by further resources.

Robeyns is well aware that individuals vary in their capacity and inclination to convert resources into flourishing, and therefore some might, with the same resources, fail to reach the satiation of flourishing. And while personal characteristics are somewhat taken into consideration the threshold is not individually tailored, therefore some people will not achieve full flourishing. To solve this problem, instead of requiring the actual realization of full flourishing, Robeyns focuses on the resources needed for flourishing, conceding that people who reach the threshold may not have obtained full flourishing. In this case, surplus resources would not have zero moral weight, because allowing the person to keep them can enhance that person’s flourishing.

Further, even if people are fully flourishing, this does not yet entail that resources above satiation have ‘zero moral weight’. Since her approach is not a hedonistic or a desire fulfillment approach, Robeyns acknowledges that people can lead fully flourishing lives and still have unsatisfied desires that the resources taken from them might have fulfilled (Robeyns, Reference Robeyns2022, 254). However, she does not explain why desires cannot have some moral weight, perhaps such that is usually outweighed by other considerations (Timmer Reference Timmer2021a, Reference Timmer2021b, Reference Timmer2021c). If, after achieving perfect distributive justice, we could choose between a world in which desires are satisfied and a world in which they are not (assuming all other things are equal), would not a world in which more desires are satisfied be a better one?Footnote ⁹ There may be additional values apart from desire satisfaction that have moral weight under various conceptions of the good life and are unaccounted for by Robeyns’ standard of flourishing satiation. Thus, for example, beauty or knowledge can, according to various conceptions of the good, be valuable without necessarily improving human flourishing.

Once again, the point here is not so much the moral value of surplus resources (although I think they can indeed have moral weight), but rather the fact that a relative threshold does not require determining one way or another. The main value judgment that relative thresholds rely upon is that resources have instrumental value to individuals with diverse projects and conceptions of the good life. This is less robust and contested and therefore more inclusive than Robeyns’ approach.

To conclude, relative thresholds are compatible with a wide range of normative commitments and therefore provide an attractive account of limitarianism, which is a partial theory of justice.

6. The limitarian intuition is a relative one

My final argument in support of relative thresholds is that the intuition that lies at the basis of the notion that it is immoral to have too much, as a political ideal as well as a moral vice, is best understood as a relative intuition. Admittedly, this does not entail that a satiation flourishing threshold is not a workable threshold, but it does support the view that whatever our approaches to distributive justice more generally, our intuitions concerning the extremely rich are relative.

An 1804 painting by Georg Emanuel Opiz, called “The Glutton” depicts a man sitting and dining. The man, an extremely fat nobleman, is sitting at a table overflowing with food. The man’s plate is soiled with the remainder of his meal, his knife and fork crossed. The man is leaning back, exhausted, evidently full to the point of explosion. His situation is accentuated by another man standing on his left, who is holding his wrist, and checking his pulse. All of this does not stop the man from sending his free hand toward a cookie tray held by a slender waiter. The painting is intentionally grotesque; Opiz aims to arouse in his audience contempt toward the eater. What is it that underlies the moral condemnation we so readily feel toward the glutton (putting aside deeply rooted prejudice toward the overweight)? Is it the vulgarity of eating so much, the lack of refinement and self-discipline? Or perhaps what is worthy of rebuke is the utter disregard the glutton displays toward the waste of food that could benefit others? The first interpretation condemns the eater for his vice but does not generate demands for justice toward other people. The second is a political wrong: his preposterous behavior constitutes an injustice toward other people. We might think that the first intuition (it is morally impermissible to have excess) is an “absolute,” non-relative limitarian intuition whereas the second (it is morally impermissible to have excess when others have too little) is a “relative” intuition. I argue, instead, that both intuitions are essentially relative.

Our disgust with gluttony, and more generally with the extremely rich, is at least partly connected, I think, to the notion of waste. While we, no doubt, judge the glutton for the lack of character that is expressed in excessive consumption, his blameworthiness is accentuated by the fact that lack of restraint is shown toward what could be used for better purposes. Imagine a person who is so rich that his pet’s food and water plates are made of gold, and his collar is studded with rubies. The distaste this elicits in us is best understood as a response to the wasteful use of something that could be put to better use. If gold was so common that people threw it out without hesitation, there would be no reason to frown upon using it for pets’ plates, or any other use.

Scarcity and shortage are an inherent characteristic of the human condition. Hume famously said that in a world without shortage the “virtue of justice would never once have been dreamed of” (Hume, Reference Hume1751, 1771, S.3.3 SBN 183–184). But even in times of abundance, there is shortage, as needs change and grow to fit the increasing available resources. It does not seem that justice will become irrelevant any time soon, no matter how wealthy our societies become, since scarcity is also socially generated (Hirsch, Reference Hirsch1976). In addition to being a precondition for the notion of (in)justice, shortage also defines what is wasteful. Throwing out food (or consuming too much of it) is wasteful because there are people who need food and do not have it. We do not chastise people for collecting too many blades of grass or stones; we do not feel that the excess being hoarded expresses a lack of refinement and self-control. There is no shortage of stones and grass blades, therefore people can do as they like with them, collect or dispose of them. The meaning of ‘having too much’ is, therefore, constituted within a relative context. It is immoral “to have too much” because resources are inherently scarce, and there are always people who desperately need them.

Two possible objections may be raised to the argument that limitarianism should be thought of in relative terms. The first is the conflation objection, and the second is the distinction objection.

The first objection involves the claim that a relative definition of richness conflates richness with advantage. When thresholds are determined relatively, it is argued, that there is always someone who is rich, even in a society where the best off have barely enough to survive; and there is always someone poor, even when all members of society have enough to lead fully flourishing lives. While egalitarians are not troubled by this objection, since they insist that redistribution is still justified (Halldenius, Reference Halldenius2022), the objection might trouble a relative-threshold limitarian, whose interest lies specifically in restricting the accumulation of extreme wealth, and not in promoting material equality in general. This may be especially troubling if the result is applying limitarian policies in a poor society, where even the relatively well-off do not have enough to flourish (Robeyns, Reference Robeyns, Knight and Schwartzberg2017, 17).

This objection is more troubling, I think, in theory than in practice. Since limitarianism addresses only extreme material inequalities, someone who is so significantly better off than others and still very poor would be rare.

Admittedly, though, there may be cases in which relative threshold limitarianism would entail taxing those who are significantly better off but are not fully flourishing. Think of a poor society, in which the despondent are barely surviving on 5 cents a day. A man who has 100 times that sum, and 5 dollars a day (150 dollars per month) still does not have enough to fully flourish, and perhaps even cannot satisfy all of their basic needs. Nonetheless, a relative limitarian threshold set at 1:100 would render this man “rich,” and any resources that exceed the threshold would be taxed.

If all members of society, including the rich, are literally starving, and survival depends on keeping any resources one can find, I concede that there is no limitarian duty to redistribute resources. But in such cases, as Hume observes, the “laws of justice are suspended […] and give place to the stronger motives of necessity and self-preservation” (Hume, Reference Hume1751, 1771, S.3.8 SBN 186–7). There is no point in redistributing resources when all are doomed.

But in other, less extreme cases of poor societies, I am willing to bite this bullet and insist that limitarian duties apply even when the rich themselves are not fully flourishing. In our example, if indeed 150 dollars a month allows more than bare survival, and considering how poor and despondent everyone else is, I would insist that having an income that exceeds 150 dollars is indeed morally objectionable. Given the contextual nature of needs and the local value of money, being so much richer than most of the other people in society almost always means having secure access to basic needs. And once sustenance is ensured, people who are so significantly better off than others in their community are under a duty of (limitarian) justice to share with others what (little) they have.

Taxing resources although they can contribute to flourishing is not especially morally problematic. First, recall, that I argued in part 4 that resources can have moral weight even if they cannot contribute to flourishing (they may satisfy desires or realize other values), so there is arguably nothing special about such taxing. But even if that were not the case, other theories of distributive justice routinely require taking resources from the advantaged and redistributing them to the disadvantaged, even when the former can still benefit from them. The benefit to the disadvantaged outweighs the value to the well-off. So taxation, and even full taxation of wealth does not rely on those resources having no value for those holding them.

The conflation objection fails also in rich societies with large inequalities. As argued in Part 2, wealthy societies in which everyone is objectively well off can still foster material inequalities that trigger the kind of political inequality that justifies limitarianism. Moreover, since the standard of living affects what needs are considered urgent, stark inequalities result in the denial of access to goods that are widely available, eventually creating deprivation that morally grounds limitarianism. Therefore, if the wealthy are more than, say, 100 times better off than the least well off, limiting their wealth may be justified, even if the disadvantaged are not poor, and may even be fully flourishing (Neuhäuser, Reference Neuhäuser and Robeyns2023).

The second objection states that when understood in relative terms, limitarianism loses its distinction, and is, essentially, reduceable to an egalitarian principle of justice. In other words, since egalitarian theories of justice oppose inequality, especially large inequalities, the kind of limitations on extreme wealth would almost always be covered by principles of equality. So while relative thresholds may save limitarianism from various objections, it arguably does so at the cost of rendering the principle obsolete.

The question of whether limitarianism is a distinct theory of justice has occupied the theory’s critics and proponents since the start, and its exact place within theories of justice is still very much on the table (Timmer, Reference Timmer2021b; Huseby, Reference Huseby2022; Robeyns, Reference Robeyns2022). As described in part 4, limitarianism is not typically thought of as a comprehensive theory of justice, but as a part of a set of principles that work together to promote justice. Still, arguably, if it is to be of theoretical and practical interest, limitarianism must contribute something to the discussion that cannot be gleaned from other principles of justice. This has been put to question, most trenchantly by Robert Huseby (Reference Huseby2022), who argues that limitarianism as presented by Robeyns, is “best seen as a combination of instrumental egalitarianism and sufficientarianism” (230).

It should first be stressed that while egalitarians advocate for equality, limitarians are uninterested in equality in general and allow rather extreme inequality to obtain before interfering. Still, one could argue that egalitarianism is a wider theory which includes relative threshold limitarianism as well as concern for other types of inequality. Seeing that egalitarianism would not allow any individual to be better off than another, endorsing egalitarianism might make limitarianism redundant, offering no distinct normative input. But this, I argue, is too quick.

First, and just to direct the criticism to the right place, not all limitarian thresholds must necessarily be relative. As I argue in this paper, limitarianism is not a unified principle, but rather a family of principles with different justifications and different thresholds that follow. Some of these, which I have not discussed in the paper, for instance, an environmental justification, may warrant an absolute rather than relative threshold, and would not be susceptible to the criticism that they are indistinguishable from a principle of equality.

But back to the point, I acknowledge that fully realized egalitarianism leaves no need for relative threshold limitarianism. In a fully equal society, there are no poor and wealthy people, let alone anyone extremely rich. But in societies in which there are inequalities (either because egalitarianism is not fully realized, or because we do not endorse egalitarian justice), limitarianism offers unique moral input, as I now detail.

Egalitarians espouse equal distributions, however when complete equality has yet to obtain, we must decide how to prioritize different inequalities and to whom redistribute resources. Relative threshold limitarianism points the finger to one particular and arguably especially morally problematic case of social inequality, offering guidance on how to respond to extreme wealth. This characterization of limitarianism, as a mid-level, policy guiding principle, has been persuasively argued for by defenders of limitarianism in previous work (Timmer, Reference Timmer2021b; Robeyns, Reference Robeyns2022). And while some may find the characterization as a policy-guiding principle underwhelming, it nevertheless demonstrates that limitarianism provides distinct and systematic answers to moral questions, distinguishing it from other theories of justice.

There is another important way in which relative threshold limitarianism is distinct from egalitarianism. As described above, the objection denotes that limitarianism is redundant because egalitarianism opposes all types of inequality, including the inequality that limitarianism targets. But relative threshold limitarians need not be egalitarians at all. Limitarianism is perfectly consistent with accepting some inequality (but not extreme inequality) as just. Thus, utlitiarians, sufficientarians, and prioritarians who reject egalitarianism as a comprehensive principle of justice, may still think that excessive wealth is unjust, for the reasons that Robeyns lays forth. If I have succeeded in making the case that relative thresholds make for the most attractive account of limitarianism, it follows that it offers unique normative considerations when applied in tandem with non-egalitarian theories of distributive justice, and is therefore not redundant.

7. Concluding remarks: defining the relative threshold

This paper argues that limitarian thresholds should be determined in light of the principle’s justifications, and that the two justifications offered by Robeyns for limitarianism are best served by relative thresholds. In addition to the fact that relative thresholds align the justifications, relative thresholds maintain the rich’s incentives to produce, they do not require committing to contested assumptions, and are compatible with the intuition that lies at the heart of limitarianism.

In this final part, I would like to touch upon some issues concerning the design of a limitarian relative threshold. Although a thorough discussion of this exceeds the scope of this paper and should be at the center of a separate study, I wish to briefly address two issues—the various possible types of relative thresholds; and how large a gap we should allow between the rich and the poor.

Before addressing these issues I should stress that not all the details concerning the desirable threshold should (or can) be supplied by the philosopher. A lot depends on facts and social contingencies, and the desirable threshold may differ between societies. The most reliable way to design a threshold to realize the justifications is to model the outcomes of various alternative thresholds using actual data concerning specific societies, a task for social scientists and economists. The justifications for limitarianism should help us evaluate the alternatives. So while the theoretical analysis should inform the design of the threshold, it does not purport to be the end of the discussion.

There are at least three possible ways to design a relative threshold: according to the size of the gap, the share of societal resources that the wealthy dominate, and the ratio between how much the rich and the poor have. I examine these three possibilities and then discuss the second issue, namely how high the threshold should be.

Size of the gap. This is the simplest form of a relative threshold, providing, for example, that it is impermissible to earn above two million dollars more than the least wealthy member’s income.

In order to determine the appropriate size of the gap, however (two million? maybe five?), we tend to think relatively. We ask, for example, what is the average income in society, or how much the most and least well-off members earn, and this helps us get an idea of the morally legitimate gap. This suggests that the size of the gap itself is an insufficient indicator of the desirable threshold, and we should rely on a more foundational criterion.

Share of all resources. Another possible way to determine a relative threshold is to limit the share of societal wealth that the rich may own. Thus, data showing that the top 1% of households own 43% of global wealth and the top 0.1% own 25% of global wealth, whereas the bottom 50% own only 1% of global wealth (TRT World, 2020), can be taken as a mark of injustice.

On the one hand, Share seems a suitable threshold for limitarianism as it is uniquely directed at extremely wealthy people who accumulate large shares of society’s wealth rather than merely well-off people. On the other hand, Share does not always result in limitarian taxing of the ultra-rich. In a society with a large middle class that holds a sizeable share of society’s wealth, the share of wealth that an extremely rich person holds may not be large enough to be taxed.

Ratio. A third standard for determining the relative threshold involves limiting the ratio between the resources available to the richest and to other members of society (primarily the poorest, but perhaps other points of comparison should be checked too). For example, limitarianism might state that the rich are not allowed to have more than 20 times the resources that the poorest member of society has, or more than 10 times the average income. Ali and Caranti (Reference Ali and Caranti2021) argue for example, that the suitable ratio for maintaining political equality in contemporary America is 1:29 (in terms of income).

Ratio applies limitarianism in more cases than Share does. As opposed to Share, Ratio requires redistribution of wealth for limitarian reasons when the poor are especially badly off compared to the richest, even if most of society’s resources are centered in a large middle class. The inverse, namely cases in which Share limits wealth and Ratio does not, can happen in principle, when a lot of society’s wealth is concentrated in the hands of the rich, but the poor are relatively well off (so the gap between them is not large enough to trigger the Ratio threshold) but seems unlikely in practice. For a small group of people to have such a large share of society’s wealth, they need to have many times more than most members of society, and that would be difficult with a large middle class, or when the poor are relatively well off.

Despite their differences, both Share and Ratio seem promising thresholds for limiratianism. The final choice of threshold requires figuring out which threshold would, in fact, best serve the justifications, considering the actual circumstances of a given society. This cannot be determined without modeling the different thresholds, and figuring out for each alternative who would be taxed, what revenues would be collected, and how wealth would be distributed in society as a result.

The last point concerns determining how high the threshold (whichever type we choose) should be. Thresholds can be set at different levels. They can be very demanding, allowing only small inequalities, or they can be set very high, targeting only the most extreme wealth.

The appropriate level of the threshold depends on what is best suited to the justification of limitarianism and therefore must be examined separately for each justification in each society. The models must rely on precise empirical data concerning possible thresholds and their outcomes, according to demographics, social wealth, other available redistributive measures, and more. Given that limitarianism is a partial principle of justice, if the models provide more than one suitable threshold, we should prefer a high threshold that limits only the most extreme wealth over a low threshold that limits wealth more widely.

Like the choice between a relative and an absolute threshold, which was at the heart of this article, the detailed design of limitarian thresholds must be informed by limitarianism’s justifications and characteristics. While I focused on relative thresholds, it is my hope that the arguments put forward in this article will contribute nuance to the discussion of all limitarian thresholds—relative and absolute alike.