Equality, efficiency and hierarchy in the workplace

Abstract

Relational egalitarians argue that workplace hierarchy is wrong or unjust. However, even if workplace hierarchy is morally deficient in one respect, the efficiency of hierarchical cooperation might vindicate hierarchy. This paper assesses the extent to which relational egalitarians must make concessions to workplace hierarchy for the sake of efficiency. I argue that considerations of hierarchy provide egalitarians with reasons that make workplace hierarchy tolerable despite being unjustified, and, moreover, that under a predominantly hierarchical status quo, the practical import of egalitarian reasons is unlikely to be undercut. This can be the case even if social hierarchy sometimes constitutes social cooperation.

1. Introduction

For egalitarians, the professional has become political. A growing body of thought holds that justice (or, more generally, morality) requires that persons stand as equals within certain social relationships (for classic developments, see Anderson 1999; Scheffler 2003). A commitment to relational equality excludes (among other things) hierarchical social arrangements, such as those exemplified by caste or feudal societies (Kolodny 2023: 90–91), and, positively, requires (among other things) that political relations be structured democratically (Viehoff 2014; Motchoulski 2021; Kolodny 2023: 323–344). This view is naturally at odds with the prevailing way of organizing firms, in which hierarchy is ubiquitous and pervasive. Consequently, relational egalitarians criticize workplace hierarchy and argue that firms ought to be organized in more egalitarian, democratic ways (Hsieh 2008; Landemore and Ferreras 2016; Anderson 2017; Tsuruda 2020; for a review of the literature see Frega et al. 2019).

This critique of workplace hierarchy raises an issue for relational egalitarians that has long been familiar to distributive egalitarians, namely, that bringing about equality will come at some cost to the well-being of persons. For distributive egalitarians, this problem is forcefully stated in terms of the levelling down objection, according to which such egalitarians are committed to holding that it is better in at least one respect to make some persons worse-off at no benefit to others when doing so reduces inequality (Parfit 1997). Distributive egalitarians have been the traditional foil for relational egalitarians, as the latter distinguish their view from distributive views by insisting that the object of justice (or egalitarian morality more generally) is not the distribution of some discrete good among persons, but rather the structure of social relationships in which they stand. This focus on relationships seemed to allow relational egalitarians to sidestep worries about the costs to individual well-being that pursuing equality may bring about. The growing discussion about hierarchy in the workplace, however, suggests that countervailing considerations of well-being cannot be so easily avoided on a relational egalitarian view. Firms are by and large organized hierarchically because hierarchy is efficient – hierarchical productive relations yield a greater supply of goods holding fixed available resources and technology. Such setbacks to efficiency amount to setbacks to the well-being of some members of the relevant society. It thus seems that relational egalitarians must reckon with such costs to efficiency.

My aim in this paper is to assess the extent to which setbacks to efficiency that arise from relatively egalitarian organizations of the workplace provide relational egalitarians with reasons to abstain from promoting relational equality. I will argue that such considerations of efficiency serve to provide reasons to tolerate some degree of inequality within firms, but that such reasons do not undercut egalitarian reasons to promote equality in the workplace relative to the status quo.

Section 2 discusses the relational egalitarian view of workplace hierarchy and argues that relational egalitarians of any bent are committed to reforming workplace hierarchy in a way that comes as some setback to efficiency. Section 3 elaborates on the idea of reasons for tolerating unjustifiable inequalities. Section 4 then considers and responds to the argument that setbacks to efficiency from undoing hierarchy are defeaters for reasons to promote equality in the workplace. My argument, roughly, is that given a predominantly hierarchical status quo, the moral importance of promoting relational equality will be greater than that of retaining marginally greater degrees of efficiency. Then, Section 5 develops a view according to which hierarchical social relations constitute social cooperation. Section 6 relies on that view to develop an argument which holds that since hierarchy constitutes social cooperation, relational egalitarians have conclusive reason to abstain from promoting equality in the workplace to any extent. Section 7 replies, conceding that when hierarchy is the sole determinant of social order then the preceding argument is sound, while arguing that in actual social settings the causes of social order overdetermine the existence of social order, such that advancing relational equality within the workplace will only come at marginal setbacks to efficiency.

2. Relational Equality in the Workplace

Relational egalitarians hold that workplace hierarchy is unjustified. For the sake of argument, I will take for granted the relational egalitarian view, and so I will not examine further the grounds for it. My aim is to assess the extent to which another good closely related to the organization of firms – efficiency – is a source of reasons for relational egalitarians to tolerate workplace hierarchy. Of course, relational egalitarians disagree on various details and subtleties about their views. So, there may be ways in which relational egalitarians can continue to disagree about particulars regarding the egalitarian organization of workplaces despite agreement on the general claim that workplace hierarchy is unjustified. Agreement on this general claim obtains on account of a growing consensus on three dimensions of social relations that matter for relational equality: ¹ (i) authority and de facto power, or the extent to which persons have formal or informal influence over decisions (Anderson 2017; Kolodny 2023: 91–95); (ii) esteem and status, or the ways in which persons are valued, respected and more generally provided with social recognition in the form of courtesy, expressions of respect, and the like (Fourie 2012; Viehoff 2019; Motchoulski 2021); and (iii) consideration of interest, or the way in which the interests of members of the relevant group are accounted for in deliberation and decision-making (Scheffler 2015; Lippert-Rasmussen 2018: Ch. 3). Relational egalitarians thus take their view to require that firms be organized in such a way that these three dimensions of social relations are sufficiently egalitarian.

Assessing the extent to which efficiency is set back in achieving equality along (i)–(iii) in firms will depend on what exactly relational egalitarians think social relations should be like within the firm. Here, it seems that disagreement among the particulars of relational egalitarian views will matter. ² It may be thought that on some accounts of relational egalitarianism, achieving relational equality within the firm comes at little to no cost to efficiency. On this kind of view, relational equality in the firm is achieved by implementing democratic corporate governance within the firm. Labourers within the firm can have an equal say with respect to capital, ownership and executives in matters of corporate governance and still choose to implement and empower efficiency-promoting hierarchical structures.

On a second construal, the demands of relational equality may be thought to extend to interpersonal relations within the firm more generally, requiring a greater degree of equality at many more stages of production. The ideal of workplace equality on this second kind of view is a holacratic one, where members of the firm self-organize and participate as equals in decision-making at many levels. So, it seems that there are two ways of conceiving of a relationally egalitarian firm. On the governance view, relational equality requires democratic governance at the highest level, but allows for extensive hierarchy below it. On the interpersonal view, firms should be more egalitarian throughout, meaning that decision-making at some levels lower than executive governance should be democratic as well.

It might then be argued that concerns about efficiency are moot for the governance view because it allows for many of the paradigmatically efficiency-enhancing hierarchical structures that are predominant in contemporary firms. By contrast, the interpersonal view makes demands for significant reforms that appear to require far more extensive setbacks to efficiency. Worries about efficiency are therefore at most a concern for only one kind of relational egalitarian view, and perhaps this is a reason against the interpersonal view and in favour of the governance view.

I doubt that so stark a distinction between the demands of the governance view and the interpersonal view can be maintained, such that concerns about efficiency cannot be an issue only for the latter. First, workers will have interests that extend beyond the scope of corporate policy. Corporate policy, by which I mean the general rules and decisions that determine the overarching aims of a firm, certainly bears on the interests of workers, as the governance of the firm shapes its ability to pay wages or provide benefits. But workers also spend a good deal of their lives within the firm. Matters such as the scheduling of breaks, availability after hours, division of assignments and organization of task groups, and hiring decisions at the lower level are all matters in which workers have interests. Even if worker interests in such matters are less urgent than their interests in corporate policy, the frequency with which a worker’s life is shaped by the mundane rules of corporate life makes it the case, I submit, that the aggregate interest over time in mundane decisions is significant. A person who spends a significant fraction of her day almost every day subject to rules that govern mundane matters of corporate life will have an interest in the organization of decision-making that determines such rules. For this reason, relational egalitarians should hold that equal consideration of interests will require equality of influence with respect to decisions about mundane matters that are beyond the scope of corporate governance.

Secondly, since relational egalitarians aim to promote equality in the social relations that each worker within a firm stands in, they must take into consideration the fact that workers are not a homogeneous group. I do not mean just that workers may disagree about the direction of corporate governance and that some structure for collective decision-making among workers will be needed. Democratic corporate governance of the firm may very well accommodate this kind of disagreement. In addition to this kind of heterogeneity of opinions about governance, workers will also stand in a variety of social relations with respect to one another. Disparities within such social relations are potential sites of objectionable relational inequality just as much as disparities between workers and executives or owners. Such disparities can matter for their own sake because they constitute objectionable relational inequality along one of the dimensions of influence, status or consideration, and they can be objectionable derivatively by undercutting equal considerations at the level of corporate policy. Workers can be exploited, dominated, and their interests disregarded by low-level management and peers just as much as they can by executives and owners. Moreover, from the fact that workers are collectively given equal influence over corporate policy with respect to executives or owners, it does not follow that each worker’s interests are appropriately considered on all matters in which worker interests should be considered. Differences in influence, status and consideration that are likely to obtain among workers will also lead to differences in consideration of interests in the collective decision-making process that is meant to convey workers’ interests with respect to corporate policy. Surely it is small achievement of relational equality, if any, to allow the most powerful or influential among workers to join the ranks of owners and executives while maintaining the rest of the status quo.

So, because workers have interests in the mundane daily organization of a firm, and because relations between workers can be the source of objectionable inequality just as much as relations between workers and executives or owners, I believe that a strict form of the governance view cannot be maintained. That is, it cannot be that the ideal of relational equality is satisfied if the structure of firms is kept by and large the way it is now in most firms, with the addition of relatively greater degree of worker influence over executive governance. Achieving relational equality in the social relations that constitute a firm will require equal influence at both the level of corporate policy as well as at least some lower-order decision-making.

To be clear, none of the above suggests that it cannot be the case that equal influence at the level of corporate policy is not the focal concern of relational egalitarianism within the firm. Moreover, I wish to be clear that the above points are not arguments against the governance view tout court. They are only arguments that the governance view cannot entirely sidestep concerns about efficiency. I think there is an appealing way in which the governance view can address these concerns. The preceding suggestions illustrate that the governance view may need to include top-down mechanisms for addressing intra-worker relational inequality or for shaping the organization of mundane daily rules in a way that brings about equal influence, status and consideration of worker interest. I think such proposals would do well to address the points raised above, but they will involve extending the scope of collective decision-making in a way that will come at some cost to the efficiency of the operation of firms. The more numbers of decisions that must be made, the greater number of considerations that bear on it, and the greater number of activities that the deciding body must carry out will all be factors that increase the costs of governing a firm vis-à-vis the prevailing, hierarchical organization of the firm (Buchanan and Tullock 1999[1962]: Chs 6 and 7). Whatever one’s preferred egalitarian conception of cooperative relations may be, one will need to take into consideration potential costs to the efficiency of production that greater equality will bring with it.

3. Tolerating Workplace Hierarchy

I began the previous section by noting that relational egalitarians take the prevailing organization of contemporary firms to be unjustified. More precisely, relational egalitarians take such hierarchy to be basically unjustified, where basic justification consists in providing a ground for the normative status of an entity based in reasons that relate to basic principles or values. It is important to be clear what kinds of reasons considerations of efficiency provide for relational egalitarians. The fact that firms will likely be less efficient if organized in a more egalitarian way is a reason for relational egalitarians to tolerate workplace hierarchy, but it is not a reason that bears on basic justification. A challenge to the basic justification of relational egalitarians’ assessment of hierarchy in the workplace would need to be a challenge to the evaluative judgements that distinguish relational egalitarianism as a normative view. That is, such a challenge would need to consist of an argument the conclusion of which would show either that relational equality is not of value, or insofar as it is, that it is of subordinate value to other considerations that are decisive in the assessment of workplace organization. A challenge to the basic justification of relational equality thus amounts to a kind of scepticism about the value of relational equality.

Appeal to considerations of efficiency as reasons for tolerating workplace hierarchy is in one respect more compelling than challenges to basic justification. Arguments that aim at toleration make the substantive concession that relational equality is of significant value, but then maintain that despite this fact, persons who believe hierarchy in the workplace to be basically unjustified will still have conclusive reason to not act in ways that would promote equality in firms. Reasons of toleration create separation between our basic evaluations and their import in our practical action. So, while an argument for toleration of hierarchy is concessive in its basic normative premises to relational equality, this concession makes the argument a powerful challenge for the practical consequences of the relational egalitarian view.

Relational egalitarians have good reason to take such a challenge seriously, given that many relational egalitarians take it to be a virtue of their view that it has a pragmatic bent. Elizabeth Anderson argues that according to relational egalitarianism, the ‘proper negative aim of egalitarian justice is … to end oppression, which by definition is socially imposed’, and the ‘proper positive aim is … to create a community in which people stand in relations of equality to others’ (Anderson 1999: 288–289). Samuel Scheffler similarly argues that the distinguishing feature of relational egalitarianism (vis-à-vis distributive egalitarian views) is that, on the relational view, ‘equality so understood is opposed … to oppression, to heritable hierarchies of social status, to ideas of caste, to class privilege and the rigid stratification of classes’ (Scheffler 2003: 22). Relational egalitarians motivate their view by arguing that it appropriately locates objectionable social inequalities as they typically obtain within the social contexts that persons live in. This provides a compelling reason to pursue reform in the social world we live in. An argument from toleration undercuts this pragmatic aspiration. For this reason, the argument for tolerating workplace hierarchy is forceful in its own right, and it is this kind of challenge that I will be concerned with in this paper.

My focus, then, is on whether, and why, considerations of efficiency are reasons for relational egalitarians to tolerate hierarchy in the workplace. Toleration is a complex notion, and I will not be able to do it full justice here. It is complex in part because it is ambivalent; the things we tolerate are objects of some kind of disapprobation on the one hand, but also objects deserving some kind of respect on the other (Williams 1996). For my purposes, the relevant features of reasons for toleration is the import of those reasons in practical action. A general feature of reasons for toleration is that they are reasons to not interfere with something that has features that we otherwise think permit some degree of interference. So, I will rely on the following account of what reasons of toleration are reasons for. Toleration concerns behaviours or states of affairs which one (a) has compelling reason to regard as normatively deficient in some respect, but (b) are nevertheless not liable to intrusive inter-reference. I might regard your religious doctrine as unjustified, and indeed as a profound threat to your well-being and flourishing, and yet despite these stakes I may not interfere with your religious practices because I ought to tolerate them.

So, arguments for the toleration of workplace hierarchy are arguments that aim to provide relational egalitarians with reasons to abstain from taking action to reform or change the hierarchical organization of firms despite the fact that such egalitarians take hierarchy to be basically unjustified and despite the fact that such egalitarians take their view to be distinctively pragmatic in guiding action toward the undoing of unjustified hierarchy.

4. Equality, Efficiency and Trade-offs

The standard case for the efficiency of workplace hierarchy draws on insights into economic organization developed by Ronald Coase (1937; cf. Williamson 1973). Hierarchies are understood to be a natural development in market societies because they make cooperation more efficient on the margins. It is, after all, surprising that in a free market, within which decentralized exchange is an engine of efficient cooperation, we nevertheless find hierarchical firms to be almost ubiquitous. Coase’s explanation for this turns on the idea of transaction costs. These are costs that obtain in the real world (as opposed to ideal models of economies) by virtue of individuals’ epistemic limitations that are impediments to successful cooperation. Examples of transaction costs are things such as the time and effort one must expend in finding an opportunity for exchange, or the time and effort spent trying to coordinate and arrive at an agreement in the terms of exchange. Coase’s proposal is that hierarchical arrangements in the workplace serve to reduce transaction costs by centralizing decision-making, thus avoiding the barriers that might obtain if the firm were to instead operate as a series of bilateral exchanges among independent workers. The fact that hierarchies make cooperation marginally more efficient explains their prevalence in the market. Competition between firms creates a selection pressure for firms that are more efficient, such that the marginal gains in efficiency afforded by hierarchical organization will, over time, lead to most firms being hierarchical.

It may be thought that there is reason for egalitarians to tolerate hierarchy in the workplace by virtue of its efficiency-enhancing property. Since hierarchy improves efficiency, and since efficiency is valuable, we have reason to maintain hierarchy in the workplace that defeats the relational egalitarian reason against it (Cowen 2017; for related discussion in the context of legal regulation, see González-Ricoy 2022: 117–118). Call this the marginal efficiency argument for workplace hierarchy.

The first important claim of the marginal efficiency argument is that efficiency is valuable. I will take this for granted. This need not mean that efficiency is of intrinsic value. Insofar as efficiency entails that some persons are able to consume a greater number of resources or services that contribute to their well-being, then efficiency is at least of derivative value by virtue of being a source of greater individual well-being.

This leaves the second key claim of the marginal efficiency argument, which is that the value of greater marginal efficiency is a conclusive reason for egalitarians to tolerate hierarchy in the workplace. I believe that egalitarians may offer a qualified denial of this claim. According to what I will call the trade-off reply, egalitarians may hold that it is sometimes reasonable to trade off some degree of efficiency for gains in equality. Egalitarians may acknowledge the instrumental value of efficiency but maintain that some setback to efficiency can be worthwhile when it comes at a great enough gain to equality. ³ The result of the trade-off reply is that egalitarians will need to tolerate some degree of hierarchy, but the degree that they have reason to tolerate is less than what exists in most firms in contemporary economies.

The extent to which the trade-off reply supports egalitarian arrangements will depend on the particular economic circumstances of a given case. This makes the present argument contextual, in the sense that the strength of reasons for or against promoting relational equality in firms is in part a function of the contingent social and economic circumstances in which we are making the relevant evaluations. Such contextualism is important, as without it I think egalitarians would rightly be accused of having a rather naïve view about hierarchy, to wit, that there should be no hierarchy no matter what. The general position that I develop here is that egalitarians must reckon with the fact that hierarchies are a remarkably effective and scalable way of producing goods efficiently. In reckoning with that fact, egalitarians do not need to acquiesce in any degree of hierarchy on account of its cooperative virtues. Nor do they need to oppose hierarchy come what may, as the naïve view would otherwise suggest. Instead, egalitarians should regard hierarchy as a highly effective piece of social technology the value of which depends on the circumstances in which it is used and the purpose to which it is put. In some cases, like those of collective emergency, there will be good reason for egalitarians to tolerate a great deal of hierarchy. I will return to these points again in the conclusion, for now, I just mean to highlight the contextual nature of the egalitarian grounds for tolerating workplace hierarchy. As it concerns my argument, my claim is that under prevailing economic conditions, there are two compelling reasons to which egalitarians may appeal to argue that trading off efficiency at the margins is worth the gain to equality.

First, existing economic arrangements favour hierarchy to such a great degree vis-à-vis relational equality that gains to the latter are possible at little to no cost to the former. Many firms are organized into a corporate hierarchy, with the various branches or divisions of the firm answering to a directly superior branch or division. And even among firms where organization is more complex or decentralized among branches, at the interpersonal level most employees are subject to the direct authority of some supervisor whose direction is typically overriding and conclusive. These facts suggest that there are reasons in favour of reducing hierarchy. Those reasons depend on the grounds we have for favouring marginal gains or setbacks to some value in a situation where we must make trade-offs among values. I submit that, given the prospect of trade-offs between two or more values, the following considerations bear on determining the all-things-considered importance of advancing one value over the others. First, given an opportunity to advance only one value, it matters to what extent each value would be advanced if one chose to advance that value. Second, if the pursuit of a value setbacks some other value, it matters the extent to which the pursuit of a value sets back the other. All else being equal, given two values V and T that we must make trade-offs between, the reason to promote V at some expense to T is stronger when: (i) the marginal gains to V if it is pursued will be greater than the marginal gains to T if T were pursued instead, and (ii) the marginal setbacks to T that arise because of pursuing V are not disproportionate with respect to the marginal gains to V.

My first claim, then, is that conditions (i) and (ii) are satisfied for the value of equality vis-à-vis the value of efficiency in many firms in the current economic context. Whether condition (i) obtains depends on the way in which equality within the firm is pursued. Not all egalitarian reforms will be on equal footing, given that different reforms will advance equality on the margins to different degrees, and will setback efficiency to different degrees. A defence of (i) would need to show that there are some methods for implementing greater degrees of equality within the organization that to some significant extent preserve the efficiency of cooperation. And, importantly, that defence must also show that the setback to efficiency was not disproportionate, as otherwise (ii) would fail to obtain.

I submit that there have been historical instances of achieving greater equality within the organization of firms from a relatively hierarchical status quo that did not come at a disproportionate cost to efficiency. The most prominent example is the emergence of the German codetermination system, which is a notable exception to the hierarchical status quo in firms within most contemporary economies (on the history of the emergence of codetermination see McGaughey 2016). Codetermination requires that in firms of a sufficient size, a works council of employees is elected to represent worker interests and negotiate agreements between workers and employers. Works councils have a variety of powers and oversight: they are entitled to vote on the selection of board of directors and other matters of corporate governance; they can establish internal agreements and informal rules in the company concerning worker relations; and they can form joint agreements with the company and workers’ unions. Moreover, corporate decision-making about various matters pertaining to workers, such as those concerning breaks, health and safety, hours and shifts, availability after hours, introduction of work groups, and termination of employees, among others, must be made in consultation with the works council. The relevant point about German codetermination is that the institution of codetermination advanced relational equality while not only having little to no set back to the efficiency of production, but in fact by producing a relatively greater degree of productivity (Smith 1991; Freeman and Lazear 1995; Hübler and Jirjahn 2003; Zwick 2004; FitzRoy and Kraft 2005; Renaud 2007).

It may be wondered whether condition (ii) of the trade-off reply is satisfied in this case. ⁴ Of course, if the above evidence is sound, the gains to equality achieved by codetermination are proportionate to setbacks to efficiency, but that is because codetermination seems to be a Pareto improvement among the two values. In making the case for trade-offs, we need to show not only that Pareto improvements would be justified, but that some genuine trade-offs between values would be justified. Imagine, then, the case of German codetermination, and hold every detail of the case fixed except the following: suppose that the institution of codetermination reduced efficiency such that Germany’s GDP per capita in current dollar value was reduced by exactly one dollar, and further suppose that this setback in efficiency is evenly distributed across the population. To make matters more concrete, suppose that the German codetermination system prevents, say, one hundred thousand instances of workplace domination that would otherwise have occurred. I submit that each German citizen having their purchasing power reduced by one dollar is worthwhile the gain to relational equality. If you find this comparison unconvincing, increase the number of instances of domination that are prevented by codetermination policies and reduce the cost to GDP per capita until you find that the gain to relational equality is worth the cost to GDP per capita. Insofar as there is some ratio at which you think a trade-off is sound, then relational equality and efficiency are partially comparable such that sufficient gains to relational equality are worth some setbacks to efficiency. So, even granting that greater equality within firms will not always have the efficiency-enhancing effect observed with codetermination (say, because it was a peculiar feature of the time and place in which it was implemented), the facts concerning the practice of German codetermination suffice to show that marginal gains to equality in the workplace can be quite significant while setbacks to efficiency are minimal. And, in adjusting the case to allow for trade-offs, we can maintain that the gains to the former outweigh the costs to the latter in at least some range of comparisons.

I now turn to a second reason in support of advancing relational equality on the margins at some setback to efficiency, which is that the relatively worse-off members of society will benefit from greater relational equality within the firm. Let me first explain how this reason relates to the first argument offered. There are at least two ways in which gains to the relatively worse-off may be related to the value of relational equality. First, having taken for granted a pluralistic view of value, benefit to the relatively worse-off may be an independent moral value that weighs in support of the same kinds of reforms as relational equality. Second, and more pertinent for the present discussion, relational egalitarians have argued that considerations of relational equality provide reasons for bringing about greater equality in the distribution of resources and other goods (Schemmel 2011; Elford 2017). So, benefits to the relatively worse-off may possess a derivative value with respect to relational equality, such that the relatively egalitarian distributive consequences that would be brought about by levelling (to some extent) workplace hierarchy count as further instances of advancing the value of relational equality. That is to say, the achievement of greater distributive equality may enhance the extent to which relational equality is in turn achieved, tilting the scales of trade-off comparisons further in favour of equality.

Here is why greater relational equality within the firm is a benefit to the relatively worse-off. Many persons who occupy relatively low positions in their workplace hierarchy also tend to enjoy fewer advantages, such as wealth or social status, than do persons who occupy a superior position in the hierarchy. Insofar as making workplace conditions more egalitarian makes persons who occupy relatively low positions better off, then such reform is a way of benefitting some of the relatively worse-off members of society. Such persons would benefit, first, by receiving a greater degree of control over the organization of a significant fraction of their daily life, and second, by avoiding the disvalue or risk of certain kinds of relational inequality such as domination, supplication and exploitation.

It may be replied that this argument from appeal to the relatively worse-off members of society does not succeed because the truly worst-off in society do not benefit from marginal gains in equality within the workplace but do stand to benefit from marginal gains in efficiency. ⁵ The worst-off, as opposed to relatively worse-off, are persons such the homeless or unemployed, or those who depend on redistribution for access to resources more generally. Since such persons are excluded from the workplace, marginal gains in equality are no benefit to them, whereas marginal gains to efficiency can be an important source of benefit.

In the first place, I doubt that gains to efficiency would benefit the worst-off. Such persons suffer severe deprivation of access to resources and various basic goods needed for a decent life precisely because they are excluded to the margins of, if not entirely from, economic activity. Economic exclusion makes it the case that one will not benefit from greater productivity within economic relations from which one is excluded.

Setting that thought aside, there is a deeper reason why the claim that the worst-off will not benefit from egalitarian reform fails to identify a reason for egalitarians to tolerate hierarchy in the workplace. The worst-off are already subject to an injustice which calls for rectification, and setbacks to efficiency further disadvantage the worst-off only because of the uncorrected injustice. If the injustice were corrected and the worst-off no longer excluded from economic cooperation, then they too would stand to benefit from relatively egalitarian workplace relations in the way described above. The problem is the standing injustice by virtue of which the worst-off are as badly off as they are. What egalitarians have reason to do, strictly speaking, is to rectify the injustice and advance equality in the workplace. Provided that the injustice will not be rectified, it may be the case that egalitarians should not advance equality in the workplace, but their reason for doing so will not be a reason to tolerate workplace hierarchy. Rather, it would be because the well-being of the worst-off is held hostage by virtue of an unrectified background injustice. Egalitarians have no more reason to tolerate hierarchy in this situation than a person has reason to tolerate a robber demanding her wallet. This claim turns on the fact that toleration is a justificatory relation; it obtains in part on grounds of reasons to abstain from interfering with some conduct or practice that one otherwise has some reason to interfere with. One does not have reason to tolerate the robber in the sense of having a practical consideration to abstain from interfering with the robber independent of the robber’s exercise of coercive power over one. Similarly, given the fact of injustice, egalitarians may acquiesce in the hierarchical arrangement of workplaces, not on account of a justificatory reason, but only because the standing injustice leaves them powerless to pursue the course of action that they take to be justified.

In sum, I have argued that relational egalitarians may reply to the marginal efficiency argument for workplace hierarchy by appeal to the trade-off reply. I have given two reasons in support of the trade-off reply. First, under a predominantly hierarchical status quo, the moral importance of marginal gains to equality will be greater than the moral importance of marginal losses in efficiency that pursuing the former may bring about, such that it is worthwhile to advance equality. Second, many of the relatively worse-off in society stand to benefit from greater equality in the workplace.

If the trade-off reply is sound, relational egalitarians will have reason to advance equality in firms to some extent while also retaining some reason to tolerate hierarchy. Reasons for toleration can obtain independent of background injustice, and instead might arise only because hierarchy is a piece of social technology that is highly effective at organizing efficient social cooperation. How far those reasons take egalitarians with respect to tolerating workplace hierarchy will likely depend on further social facts about hierarchy and on the finer grained details of one’s relational egalitarian view. Since the trade-off reply does not result in the conclusion that all workplace hierarchy is intolerable, it will not be the case that relational egalitarians have reason to advance equality in the workplace to the fullest extent. Insofar as individual well-being is also a value and hierarchy an efficient tool for the production of goods that advance individual well-being, egalitarians will have reason to rely on hierarchies as an instrument to advance individual well-being. This is not acquiescing in injustice, but a response to the fact that there are competing values that direct us to organize social life in different ways.

5. Hierarchy and the Constitution of Cooperation

I will shortly consider an argument that may be offered to egalitarians to tolerate hierarchy to a much greater extent. This argument turns on the claim that hierarchy is constitutive of stable cooperative social relations.

The line of reasoning that leads to the idea that hierarchies are constitutive of social cooperation shares a starting point with egalitarian condemnations of workplace hierarchy. Basic to the egalitarian critique of workplace hierarchy is recognition of the fact that workers are not mere nodes in an organizational chart but persons who will have to stand in social relations with one another while cooperating. Hierarchy in the workplace is not just a coordinating device like a traffic light on the road. Persons must stand in social relations of superiority and inferiority in order for there to be a hierarchy, and these relations among persons, egalitarians argue, come at a substantial (and unjustified) moral cost.

The case for hierarchies constituting cooperation turns on the idea that persons tend to be motivated to be superiors within social relations, such that persons in the workplace will be motivated to take on some costs in pursuit of occupying relatively superior positions.

If persons are motivated by a desire for superior status in the workplace, then there appears something puzzling about the hierarchical organization of cooperation. Why do workplaces not devolve into competitions over status? Such competition appears occasionally in workplace politics, but seems to be, for the most part, rather restrained. Given the deep positionality of one’s place in the workplace hierarchy and given that person’s desire to have relatively superior positions, we should expect conflict over hierarchical position to be prevalent. Satisfaction of one person’s desire for a superior position comes at the expense of another who will remain motivated to compete for superiority. Modern workplace hierarchies are usually highly stratified with distinctions of many different levels of superiority and inferiority. Under these conditions, persons who desire to be the social superior of others and who are willing to take on costs to do so should be caught in cycles of persistent competition over positions of superiority. Formal analysis of hierarchical cooperation comes to this conclusion, showing how cooperation can become unstable and break down when persons engage in too much competition over status (Kitts 2006). Why then, is such hierarchical conflict not observed in the firm as often as this theoretical argument suggests it should be?

To get a clearer picture of the problem, let us a consider a model of hierarchical organization which treats conflict over hierarchical position as endogenous to the process of achieving cooperation. ⁶ To keep the analysis tractable, I will assume that interactions are between two players, who I will call Player 1 and Player 2. Each has the same two strategies available to them; they can either Lead, or they can Follow, and the players will choose their strategies simultaneously. If exactly one player leads, then there is a determinate leader who will coordinate cooperation, and that leader enjoys the benefit of greater social status, represented by S. If neither Lead, then no party to this conflict will coordinate collective action. This need not mean that coordination will not be achieved. For the sake of simplicity, I will take for granted that some leader is always chosen, so that the benefits of coordination are always secured – the winner of a different leadership contest might take over, the contest might be repeated, or perhaps a member of the population is chosen at random to be the leader. Of greatest interest for us is the case where both parties choose to Lead. In this case, the players enter into conflict, and the winner of the conflict will be deemed the leader. Conflict in the model is meant to capture the familiar stuff of hierarchical politics, which involves costly activities such as coalition building, cajoling, and ingratiating, which detract from the cooperative endeavour at hand. Because such conflict is costly (time spent greasing the wheels of politics is time unspent on more productive tasks), both parties will take on some cost C (which, for simplicity, is assumed to be symmetrical), and since the outcome of such conflict is uncertain, the probability P (between 0 and 1) represents the likelihood that Player 1 wins out. The interaction just described is represented in extensive form in Figure 1.

Figure 1. Leadership game.

There are three Nash equilibria in the leadership game. The first two are pure equilibria, (Lead, Follow) and (Follow, Lead). The third is the impure equilibrium where Player 1 plays Lead with probability 1/(1 − P + C/S), and Player 2 plays Lead with complementary likelihood, 1/(P − C/S). Under the impure equilibrium, there is some probability, given by the conjunction of the two probabilities stated in the preceding sentence, that both players Lead, and so incur the cost of conflict.

As the likelihood that some player will win out in the conflict increases (as P approaches either 1 or 0), the impure equilibrium will approach one of the pure equilibria. So even if conflict is possible, its likelihood is minimized when it is clear that in the case of conflict, one particular player is more likely to win out. ⁷ I propose that social status is one tool for pre-empting conflict over hierarchical position, which is why conflict over superior positions is not as prevalent as we might otherwise expect. In established status hierarchies, it is effectively predetermined who will win out in a given case of conflict. Superiors normally trump inferiors when conflict arises, and this fact is known in advance by all parties. Conflict in cooperative settings is most commonly observed to occur in cases where status hierarchies are ill-defined (Loch et al. 2000). Moreover, the model seems to accurately capture where we should expect most conflicts over superiority to occur in firms, namely, among persons who presently have relatively similar status within the firm such that either has (what they believe to be) some significant prospect of winning out in conflict over a superior position.

The presence of well-defined status distinctions, then, can be understood as driving the value of P in the leadership game to either 0 or 1, pre-empting conflict over positions of superiority and the attending social costs. This is the sense in which hierarchy is constitutive of cooperation – hierarchy in the workplace makes cooperation possible in the first place by pre-empting costly conflict over positions of superiority which, if unchecked, may undermine cooperation.

6. The Constitutive Cooperation Argument for Tolerating Workplace Hierarchy

If it is granted that hierarchies are constitutive of cooperation, it may be argued that relational egalitarians should tolerate all degrees of workplace hierarchy in existing economic arrangements because attempts to bring about equality will result in a strictly morally worse state of affairs. This argument is similar in spirit to the marginal efficiency argument considered in section 4, though it is much stronger. The general intuition remains the same: in the absence of hierarchy, we will reap fewer goods from cooperation. The present argument is stronger in that it claims not only that hierarchy makes cooperation more marginally efficient, but further that hierarchy enables cooperation in firms as such. The thought being that in the absence of hierarchy, cooperative social relations will dissolve into conflict over social position. Further support for this argument might be found in the fact that the overwhelming majority of cooperative human enterprises, which includes much more than just firms, are, and have been, organized hierarchically.

The argument for tolerating all degrees of workplace hierarchy on grounds of it constituting cooperation is as follows. If hierarchy is constitutive of cooperation, then attempts to reform hierarchy will result in a strictly morally worse state of affairs, granting one further premise: the shared beliefs, expectations and evaluations that make up a hierarchy are resistant to change. To use a physical metaphor, social practices have some inertia, such that changes to shared practices will often be slow and incomplete (Hannan and Freeman 1984). Within the context of firms, there are at least two reasons that can serve to explain why social inertia might obtain. First, some persons will have motives of self-interest to act on and maintain the social hierarchy. Those relatively close to the top of the hierarchy benefit greatly from their position, and compliance with the norms and expectations comprising the hierarchy contributes to maintaining that beneficial position. Second, individuals may have been habituated to comply with hierarchical norms and expectations. Norm internalization is a ubiquitous feature of most any social practice. When persons internalize norms, they develop an internal source of motivation to comply with norms and come to perceive their social world as being appropriately shaped by the norms they have internalized (Bicchieri 2005: 146, 173–175, 194). Persons who have internalized hierarchical norms, then, will have robust dispositions to comply with hierarchical norms and to see them as fitting for the workplace.

The social inertia of hierarchies will result in egalitarian reforms of workplace relations producing strictly morally worse outcomes. To return to the terms of the leadership game, well-defined hierarchies the norms of which are internalized drive the impure Nash equilibrium of the leadership game to one of the pure equilibria, pre-empting conflict over status. If reform is introduced requiring relatively egalitarian arrangements within hierarchical firms, this will change, to some extent, expectations regarding who wins out in leadership conflicts because egalitarian arrangements provide their own mechanism for collective choice and determination of leadership roles. But the norms and expectations of workplace hierarchy are subject to social inertia, so we should expect such reform to have a limited effect in displacing the hierarchy. Consequently, there will still be some individuals who expect hierarchical norms to be followed, and so will expect, and support, individuals ranked higher in the traditional hierarchy to win out in the case of conflict. The resulting state of affairs is one in which some individuals will believe they have a mandate to lead on egalitarian grounds, and others will believe they have a mandate to lead on hierarchical grounds. Conflict over leadership will be more likely, as a greater number of persons will come to think that they are likely to win out in the kind of conflict modelled by the leadership game. Since the egalitarian reform will have some effect on the norms and expectations of individuals, we should also expect the probability that a given party wins out in conflict to be further from 0 or 1 than it would have been had the hierarchy been undisturbed. Reform that requires egalitarian arrangements within firms to result not in egalitarian arrangements, but rather in a relatively ill-defined hierarchy. The claim, then, is that an ill-defined hierarchy is strictly morally worse than a well-defined one because it still fails to achieve egalitarian organization while losing out on the cooperative virtues of hierarchy. When hierarchy is constitutive of cooperation, it seems that egalitarians have reason to tolerate the full extent of the hierarchy that presently obtains.

7. The Trade-off Reply Revisited

There are two ways in which the trade-off reply can be raised in response to the argument that hierarchy constitutes cooperation, the second of which will be successful. The first argues that displacing a hierarchy that is constitutive of cooperation produces a state of affairs that is better in at least one respect. The second argues that hierarchies are rarely, if ever, exhaustively constitutive of cooperation.

First, one could argue that the claim in the prior section that reform of a hierarchy that is subject to social inertia results in a strictly morally worse state of affairs is false. It may be that there is one respect in which a more ill-defined hierarchy is better than a well-defined one, namely, in that the former is more egalitarian (because less hierarchical). This gain in equality may then be argued to be worthwhile on grounds of the trade-off reply.

I do not think that this version of the trade-off reply is successful. It is not the case that since there is less hierarchy there is more equality; reform of a hierarchy that is constitutive of cooperation does not make for a more egalitarian social order, but less social order as such. This claim is based on an understanding of social order according to which constitutive of social order is the resolution of problems of discoordination and social conflict. Social disorder, then, is a state of affairs in which persons fail to coordinate on terms of social life and instead find themselves in costly conflict with one another. If it is granted that hierarchy is constitutive of cooperation, displacing hierarchy will not bring about more equal social relations but will displace ordered social relations as such.

What should be challenged instead is the claim that hierarchy is constitutive of social cooperation. Is it the case that cooperation within firms is organized solely on grounds of hierarchical social relations? If one were to accept the claim that hierarchy constitutes, in an essential way, all cooperation, one would be led to expect that any displacement or frustration to cooperative hierarchy would in turn result in social disorder. But of course this is not true. It may be the case that conflict in cooperative social settings is more prevalent when hierarchy is ill-defined, but this does not entail that every setback or frustration to a given hierarchical organization results in social disorder.

It is not the case that every disruption to hierarchical organization results in some greater degree of social disorder because of the fact that we order our social relations by relying on a variety of mechanisms, among which hierarchy is just one. More generally, the extent to which hierarchy is constitutive of stable cooperative social order should be understood in terms of thresholds. Hierarchy is needed to some extent in order for cooperation to obtain and not devolve into costly conflict over social position. But the threshold of hierarchy that is needed is a function of the other sources of social order, such that insofar as there are other determinates of stable cooperation, the extent to which hierarchy is needed to enable cooperation is lowered. My claim is that due to the array of overlapping sources of social order within cooperative settings, the threshold of hierarchy needed to enable cooperation is very low, if hierarchy is necessary at all.

The following is a non-exhaustive survey of overlapping sources of social order that organize cooperation within the firm. First, and perhaps most obviously, most persons who are part of a firm depend on successful cooperation to make a living. Since participation in a firm is the primary means by which many persons earn an income and acquire access to resources, most members of the firm will have prudential reasons of self-interest to comply with the rules and organization of production within their firm. Second, members of a firm generally become members by exercising their legal power of contract. Modern day economies are dependent on the law as an instrument of coordination, and most persons in such developed economies internalize, for one reason or another, a habit of obeying the law. This habit of obedience will carry forward to the terms of contracts that persons with such a habit make, yielding a second source of social order within firms. Third and related, persons may take themselves to have moral obligations to comply with the terms of cooperation that constitute the firm. They may conceive of their signing a contract as of making a promise or may think of themselves as having a duty to do their fair share in producing cooperative benefits.

Self-interest, habitual obedience of legal rules, and moral considerations, among perhaps other mechanisms, overdetermine the existence of social order within firms. Hierarchy is therefore one among many sources of social order, and so while it may be the case that hierarchy constitutes cooperation to some degree, it is not the case that cooperation is made possible only by virtue of hierarchy. The argument from constitutive cooperation raised in section 5 is therefore overstated. Indeed, the constitutive cooperation argument collapses into the marginal efficiency argument for hierarchy in the workplace. Since hierarchy does not exhaustively constitute cooperative social order, undoing hierarchy will not make cooperation infeasible, at most it will make cooperation marginally less efficient. Reform that moves from hierarchy to equality will at most generate greater rates of conflict at the margins within cooperative settings. Relational egalitarians may therefore answer the constitutive efficiency argument the same way that they can answer the marginal efficiency argument – by appeal to the trade-off reply. It may be that greater economic efficiency is of value, but so too is equality, and in a social environment where the former is privileged to a significant extent over the latter, we will generally have undefeated reasons to pursue the advancement of equality to at least some degree.

8. Conclusion

Modern economic arrangements rely on hierarchy in firms to organize a great deal of productive activity. Egalitarian claims to level such hierarchy must take into account the costs to cooperation that such reforms will bring about. Those costs will at most be reasons to tolerate hierarchy for egalitarians, given that costs to efficiency will ultimately bring about some lesser degree of well-being for some participants in the economy. Of course, it remains an open question to what degree hierarchy is tolerable for those who take it to be unjustified. Addressing this question will naturally require consideration of empirical details that are beyond the scope of this philosophical analysis. However, I do not think it satisfactory to leave matters there, and so in my closing remarks I would like to address the general considerations that bear on the degree of hierarchy that egalitarians have reason to tolerate.

The crux of the egalitarian trade-off reply is that the moral importance of marginal gains to efficiency will likely be less than the importance of marginally greater equality given the prevalence of hierarchy in existing economic arrangements. To assess the degree of tolerable hierarchy, we will need to consider the facts by virtue of which marginal gains to efficiency are important. The degree of hierarchy that is tolerable will be a function of the moral importance of the gains to well-being that greater efficiency would bring. Such gains to well-being can be of greater or lesser moral importance in proportion to how great the gain in well-being is, to whom the gains in well-being accrue, and how well-off those who would benefit will be otherwise. When the gains to well-being from marginally greater efficiency in production would be relatively lesser, or would accrue primarily to the relatively advantaged, only lesser degrees of hierarchy will be tolerable. Conversely, when the gains to well-being will be significant and will accrue to persons who are relatively worse-off and would be badly off without the benefit, then greater degrees of hierarchy will be tolerable.

This last point is an important one, as I think without it, relational egalitarians may be thought to be committed to an implausible view. On a crude view of relational egalitarianism, one might think that egalitarians oppose hierarchy in any circumstance, however dire the need for efficient cooperation. One might then readily misconstrue relational egalitarians as naive utopians, opposing hierarchy even in the case of social emergency, examples of which might include the production of goods for collective self-defence in a just war or the production of medicine during a public health emergency. This line of thought misconstrues the egalitarian position as being entirely insensitive to the cooperative virtues of hierarchy. Though egalitarians take hierarchy to be unjustified, they may still recognize it as tolerable on account of its cooperative virtues. Cases where the benefits of social cooperation are especially urgent, as with a just war or public health emergency, are those where the greatest degree of hierarchy will be tolerable to relational egalitarians because in such cases the amount of well-being at stake is great and many persons will be badly off if they do not receive the benefits in question. Egalitarians need not find all hierarchy uniformly intolerable, and so they may recognize exceptional cases where hierarchy should not be undone to any extent. This fact is compatible with egalitarians finding cooperative hierarchy in general unjustified, and with their finding hierarchy in the workplace to the degree that it occurs in many firms intolerable. Some degree of hierarchy within the firm will, I think, prove to be tolerable to relational egalitarians, but given the prevalence and pervasiveness of hierarchy in most modern firms, this fact does not undercut reasons to promote relational equality relative to the status quo. So, egalitarians, for the time being, need not compromise on the practical implications of their view.