Exchange and solidarity

Abstract

According to the model of exchange as mutual assistance, an exchange can be perceived as a joint activity for mutual benefit – and needn’t involve any self-directed motives at all. This essay pushes back against this new defence of market motives. The essay develops an alternative ideal of production as caring solidarity, in which production is a joint activity of caring about one another. Points of overlap and difference are developed in some detail. The essay concludes by discussing the implications for an economics of caring solidarity, with discussion of the limitations of various market socialist strategies.

1. Introduction

For as long as there have been markets, there have been disputes about market motives. For much of this history, the two sides have talked past one another. Optimists about markets have mostly addressed other optimists, and failed to take seriously the kinds of relational values that might be at stake and the range of possible alternatives to market-based production. Pessimists about markets have mostly addressed other pessimists, and failed to take seriously the full range of market-involving economic structures and the different kinds of psychologies compatible with market interactions.

It is common for pessimists about markets to assume, often without much argument, that market exchange involves objectionable egoism, or at least non-tuism – a lack of concern about the interests of the other party to a market transaction. Recent work refutes this assumption. It has been persuasively argued by Robert Sugden and Luigino Bruni (2008, 2009; Sugden 2018), and by Harrison Frye (2023), drawing on recent work on team reasoning and philosophical psychology, that market exchange might be a joint activity undertaken for mutual benefit. This mutual assistance model shares a number of the features of ideal production: participants see one another as fellow collaborators, sharing the responsibility to meet their collective needs, and might even lack self-regarding motives altogether. This presents an entirely new and sharply presented challenge to critics who think there is something objectionable in principle about exchange-based management of production.

This is the most exciting development in these debates in a long time. I agree that exchange can be perceived as mutual assistance, and I augment the view in one or two places to make it even more appealing to opponents of markets. Still, I will argue, there remain deep differences between the ideal of mutual assistance and an alternative ideal of caring solidarity. I develop this alternative ideal by drawing on recent work on care, ¹ on solidarity ² and on non-alienation. ³ As I will develop the idea, the ideal of caring solidarity is a complex of mutual care and solidarity collaboration in which we identify with one another as needing mutual care. This is a collaborative ideal, one in which individuals see themselves as participants in the collective process of doing what we can together to meet one another’s needs. This ideal has implications for the organizational structure of production that are incompatible even with exchange as mutual assistance.

Much is at stake. It is common to distinguish broadly distributive and relational objections to market-based production. Decades of important work by market socialists have provided good evidence that many distributive challenges to markets (usually different forms of inequality and exploitation) can be met, at least in principle, by suitably well-designed market structures. ⁴ But relational objections cut deeper, and threaten to impugn the basic apparatus of market exchange itself. The model of mutual assistance supports optimism that these objections, too, can be met. This essay argues that they cannot.

A clarifying word about the dialectic. I have no objection to the primary conclusion of Sugden and Bruni that exchange is compatible with the ideal of mutual assistance, in which we work together to realize mutual benefit. However, in various places they suggest a bolder conclusion, that exchange is thereby compatible with any attractive ideal of sociality in production. ⁵ Whether or not this is an official conclusion of the argument, it is a terrifically provocative conjecture. It is a challenge to market pessimists to render any residual opposition to exchange more clearly. This is the challenge I take up here.

2. The Metaphysics of Exchange and Mutual Assistance

2.1 Exchange as Mutual Assistance

Let us start by clarifying the conception of exchange as mutual assistance. It is standard to define exchange as the ‘voluntary transfer of private goods or services for consideration’. ⁶

Not all pairs of voluntary transfers of private goods or services are joint actions. I can scratch your back and you mine without these individual services being conditioned upon one another. Neither is it the case that all mutual transfers that are joint actions are exchanges. Imagine a third party puts a gun to our heads, or offers to donate money to a worthwhile charity if the two of us can work together to make one another better off in some specific respects. There is nothing conceptually impossible in our meeting this condition together. But our joint activity would, plausibly, not be an exchange of the goods that make us better off in these respects.

Exchange is a social practice distinguished by conformity to a social rule according to which the mutual transfer has a kind of reciprocal conditionality. This conditionality cannot merely be cashed out in simple dispositional terms. It may be that I am going to exchange A for B with you, but if you can’t produce B, then I’m just going to give you A. I may even be more likely to just give you A than exchange A for B with you, but we end up exchanging. A.J. Julius (2013) argued that reciprocal conditionality cannot be understood in terms of conditional intentions on the part of both parties, either. It seems that exchange itself is best understood as a joint activity that both parties participate in. Actually, an exchange is at least two joint activities, for it consists in two parties doing one thing together that is distinguished by its conformity to a social rule. The social rule itself presumably obtains in virtue of further social activity. ⁷ This is analogous to the sense in which any social game is multiple shared activities: the specific game of chess you are playing with your friend, and your adherence to the conventional rules, for instance the rules of castling and en passant.

Not all exchanges are mutually beneficial. You might agree to exchange just for my benefit, or to please a third party. An exchange might be mutually beneficial but not intentionally, even without both parties knowing that it is mutually beneficial. But a joint action might also be intentional under the description ‘an exchange for mutual benefit’. In such cases, there is another joint action taking place. In addition to the exchange, there is the doing-of-the-exchange-for-mutual-benefit. That second thing is also something participants can do together.

Importantly, it is possible that participants can do this together – can engage in a joint-activity-for-mutual-benefit – while neither party is individually motivated by mutual benefit as such. ⁸ It is possible that the two parties are extrinsically incentivized to exchange-for-mutual-benefit. To see this, imagine that two employees of two companies were each ordered by their boss to spend an enjoyable afternoon together to further the broader corporate relationship. Each employee secretly dislikes the other company. But they need the money and they might be watched, so they work together and make sure they enjoy themselves. Or perhaps we are both friends of a third party, and each of us spends time together, each trying to ensure that each of us have as much fun as possible, but just because each of us wants to please the third party. ⁹ The third parties care about the complex activity – that the joint activity produces mutual benefit – and not merely about its consequences – the mutual benefit. And the participants could themselves be motivated to participate in this more complex activity, so it seems to me, without either of them otherwise being motivated to bring about the separate state of affairs produced by the activity, namely their mutual benefit.

So, there is a distinction between mutual benefit as the end of a joint action (for instance, exchange), and mutual benefit as the motivations of the individual parties to such an exchange. The sense in which we might perceive exchange as a joint action for mutual benefit, according to Sugden, is that joint benefit might be part of the end of a joint action in this former sense.

Then there is a further question about the motivations of the participants to enter into such an exchange. Sugden says that being motivated to promote mutual benefit is a virtue (2018: 280). So, for simplicity’s sake, and to try to push as far as possible the idea of exchange as non-self-interested reciprocity, let’s presume that both parties are motivated to enter into any given exchange precisely because it will realize mutual benefit, perhaps in addition to broader background considerations about participating on market terms being a way of realizing mutual benefits across the economy more generally. Here we can point to a further argument for market-based conditionality, which is that adherence to market prices constitutes participation in a large-scale system that efficiently regulates supply and demand of privatizable resources. There are epistemic and motivational versions of this sort of argument. Let us assume, in order to continue giving the best possible case for markets, that these are good arguments (at least pro tanto), and that one might have the following layered attitude to exchange: one’s willingness to pursue and adhere to the terms of mutually beneficial exchanges – including to withdraw service if one isn’t going to get paid – is itself conditional upon one’s conviction that general adherence to the market system itself, in the relevant domain, is mutually beneficial.

2.2 Structural Overlaps between Mutual Assistance and Solidarity

This way of thinking about exchange shares with the solidarity alternative to come the following advantages over more familiar models of exchange as self-interested coordination.

Firstly, the mutual assistance approach emphasizes the importance of a ‘we-perspective’: of working together and sharing our productive responsibilities. As Sugden puts the point (2000: 182): ‘the individual frames the problem, not as ‘What should I do?’, but as ‘What should we do?’’. This approach avoids the kind of strategic thinking about the actions of others that is characteristic of game theoretic reasoning ¹⁰ and consequentialist reasoning, in favour of we-reasoning, in which one thinks in the first instance about what the group should do, and accepts responsibilities as a participant in the group. This involves a kind of trust in others. One acts on the basis of normative expectations rather than descriptive expectations about the behaviour of others. This is important. The mutual assistant is not conditioning their service on their being served, but on their both being benefitted by the relevant exchange.

Secondly, this approach avoids the simple self/other motivational dichotomy that makes much of the discussion of economic motives unhelpfully superficial, in favour of social motivations, allowing that market participation might be social ‘all the way down’. There is a sharp distinction here between the treatment of motives and welfare for the mutual assistance approach: it is compatible with rejecting any motivational distinction between self and other; however, it relies on antecedent distinct facts about what would benefit each individual.

Thirdly, this new proposal is – perhaps surprisingly – consistent with G.A. Cohen’s famous characterization of the motivational profile characteristic of the socialist. Cohen explicitly contradistinguished these motives from motives in the market. He said that the socialist finds value ‘in both parts of the conjunction – I serve you and you serve me – and in that conjunction itself: [I do] not regard the first part – I serve you – as simply a means to my real end, which is that you serve me’ (Cohen 2009: 43). The mutual assistance model is one plausible way to spell this idea out. Sugden’s virtuous market participants are motivated to pursue joint activities aimed at mutual benefit; in doing so, they plausibly value both conjuncts and the conjunction itself. They don’t need to have any attitude at all to the mere conditional, that I serve you only so that you serve me. It is quite consistent with this that these mutually assisting participants value this way of relating to one another, and engage in it (at least partly) for its own sake, as a way to sustain relationships of mutual assistance, or to use Sugden and Bruni’s term, fraternity.

Fourthly, the ideal of mutual assistance has application beyond the domain of impersonal production. Mutual assistance is an ideal of sociality that could be maintained to apply in any domain of social life. It is possible that individuals committed to mutual assistance are not self-interested in any domain of activity. Indeed, it is possible that individuals lack egoistic motives even alongside other motives, as in mixed-motives models of exchange. ¹¹ Individuals might be team players all the way down, as it were. Sugden and Bruni do not insist upon this point, but their innovation makes it an intriguing possibility.

The views being compared in this paper – the mutual assistance view and the caring solidarity view – have these structural features in common: they both involve shared reasoning, avoid the crude self/other motivational dichotomy, allow that individuals might value social production for its own sake, and offer a common ideal across impersonal production and other parts of social life. ¹² However, significant differences remain. Let me turn to develop the ideal of caring solidarity on its own terms before drawing some contrasts.

3. The Ideal of Caring Solidarity

The ideal of caring solidarity is a complex of mutual care and solidarity collaboration in which we identify with one another as needing mutual care.

It is standardly held that caring about someone involves doing what you can to meet their needs (see Noddings 1984, 16; Tronto 1994: 105; Bubeck 1995: 129; Kittay 1999; Barnes 2012: 5). This is a deep and underappreciated point of connection between care ethics and socialist ethics. The emphasis on care supports an interpretation of the famous slogan – from each according to their abilities to each according to their needs – as articulating an intentional relationship between agents: one in which an agent is robustly and intrinsically motivated to do what they can to recognizably meet some particular needs. ¹³ In being intrinsically motivated, one is motivated to meet the need for its own sake, not merely as a means to something else (such as meeting one’s own need to meet a need, though one might be motivated by that, too). This motivation is also robust, since it is not conditional on other things one wants. One’s particular motivation is not thereby unconditional, however, given the possibility of conflict with other specific commitments.

In addition to being intrinsically motivated by needs, one is also attentive to the relevant needs. Caring is a participant stance rather than observational stance, an ‘engrossment’ in the other, one that involves really listening, and taking the other’s views about their needs seriously. ¹⁴ I’ll say that this kind of humble responsiveness to the other constitutes agent respect. ¹⁵

A third relevant feature concerns the uptake of care. As Nel Noddings has put the point in a discussion of ideal caring relations (1984: 6):

How good I can be is partly a function of how you – the other – receive and respond to me. Whatever virtue I exercise is completed, fulfilled, in you.

This sensitivity to uptake involves a kind of need for recognition. This is an essential part of ‘completed’ care. This language of ‘completion’ is again suggestive of the 1840s Marx. ¹⁶ This is another point of connection between these two traditions, and, substantively, another respect in which we are vulnerable to those we care about.

Now. Caring is engrossing, but still involves a clear self-other distinction. An agent cares about a patient. In Noddings, the one-caring cares about the cared-for. In Tronto and Fisher’s four stages of care, there is a caregiver and a care-receiver. This is true even given that care involves agent respect. For A might care about B, respectfully, even though B doesn’t care about A. The cared-for may willingly volunteer authentic information about their needs, and they may accept assistance. But they may not otherwise facilitate their being cared-for. Care does not require reciprocation (to borrow language used in this context by both Marx in Comments on James Mill and by Andrea Sangiovanni (2023: 247)).

At this point, we can, like Sugden and Bruni, draw on recent work on joint activity to articulate an alternative position, appealing to the idea of caring together.

I suggest we start with the account of solidarity-as-joint-activity developed by Andrea Sangiovanni (2023). On this account, we act in solidarity when, as a result of identifying with one another, we each intend to participate to overcome significant adversity by pursuing a shared goal; we are individually committed to relevant ends and means and to ‘not bypassing each other’s will’ in pursuing them; we are committed to sharing our fate in relevant ways; and we trust one another to meet all these conditions. This is a more robust model of joint activity than Sugden’s model of team reasoning. ¹⁷ However, for current purposes, these differences are far less important than the similarities, which concern the importance of reasoning together. This involves starting with a ‘we’ perspective and consequently seeing one’s own responsibilities as a way to participate in a larger collaborative enterprise in which one trusts others to uphold their responsibilities similarly in turn.

It is important to resist three limitations on the possible extent of caring collaborations.

The first is that care is not merely a personal attitude. ¹⁸ We can care about others equally well in non-personal contexts. To see this point, it is helpful to disambiguate care and love. Plausibly, all (real) love involves care. But we often care about people that we don’t love, for instance at work, or in municipal politics. One can care about strangers, about neighbours, or about large groups. The central case of care that I’ll have in mind is vocational: you can manifest care about others as a dentist, a council worker, an engineer, a teacher, and so on.

The second point is that care needn’t be restricted to basic needs. The literature on care largely prescinds from stipulating a general account of needs, emphasizing instead the importance of responsiveness to particularity. However, care theorists sometimes assume that needs are restricted to basic needs or even to our physical and emotional needs (see Noddings 1984; Kittay 1999). This can lead to a narrow view of caring as ‘pink collar work’. Sugden and Bruni’s discussion of care work (2008: 57–63) is implicitly restricted to this sort of caring-for activity. But caring solidarity needn’t be restricted to basic needs. Our needs can be quite expansive, corresponding to the broad range of activities that can constitute our flourishing; these can include needs associated with a range of aesthetic, intellectual, and creative ideals – as well, of course, as our needs to pursue many of these ideals in caring solidarity with others.

The third point is that care is not merely dyadic. Standardly, individual caring activities are either directly collaborative or indirectly empowered by other caring participants. Here is Joan Tronto (2013: 152) on this point:

Think of some cases of supposedly dyadic care: the doctor/patient relationship, the mother/child relationship, a student/teacher relationship. Doctors do not provide health care alone; they are increasingly involved in a complex set of social relationships of care …

It is possible to see our caring relationships with particular others as participations in collaborations in increasingly expansive caring systems. ¹⁹ One sees oneself as a collaborator not merely by having a subset of the total needs that fall within one’s delegated responsibility, but as meeting those needs in collaboration with, and suitably empowered by, a range of others.

Without these limitations, we get a model of caring collaborations that can extend to a broad range of human needs, that we can aim to meet in suitably sized collaborations that are themselves collaboratively responsive to larger-scale collaborations (cf. Dewey 1888; Pateman 1970; Tronto 2013). In each of these contexts, we see ourselves as participants, sharing the project of meeting our needs together, each of us disposed to do what we can to meet needs through suitably delegated responsibilities.

We can end with a word about common identification in such a system, or how we ‘perceive’ one another in caring solidarity, to use Sugden and Bruni’s language. One thing emphasized by many writers in the literatures on care ethics and solidarity concerns the basis on which we relate to one another, which is our human vulnerability. ²⁰ We all need to be cared about as young, as old, and as infirm. And if we are to care about others we need to be empowered to do so. On this approach, to identify as humanly vulnerable is to identify as needing care and needing to care, or equivalently, to identify as needing mutual care. This is a contingent identification. Given the implicit generality of the identification as human, those who identify in this way are disposed to share relationships of this kind with anyone they identify as human, that is, anyone vulnerable in these ways, who they can trust in the relevant ways. This common identification is the last condition of the full ideal of caring solidarity.

All this has substantive implications for the nature of our joint activity of mutual care. It implies that, as participants, one of our goals is to empower one another to engage in this very activity, that is, to care about others in suitably collaborative ways. Seeing ourselves as participants in such a collaboration has implications for our willingness to take responsibility for different tasks, if doing so would better fit with some larger division of responsibilities. Again, this idea of sharing overall responsibilities is compatible with differentiation in specific responsibilities. A’s caring for B as a friend doesn’t make it appropriate for A to teach B’s students, but rather to empower B to teach B’s students. A might separately be disposed to teach B’s students if B is unable to do so. It is also plausible that, seen as an activity undertaken in light of the ideal of human solidarity, B’s end in teaching B’s students is partly characterizable as empowering those students to care about others in specific ways – as engineers, doctors, plumbers, and so on.

4. Contrasting Mutual Assistance and Solidarity

4.1 Abilities and Needs Without Mutual Benefit

Let me start with the most obvious cases, namely those in which people are unable to exchange because they cannot mutually assist one another. This might be because one has abilities without needs, or needs without abilities, or abilities employed on non-market needs.

If I am rich, I won’t have to serve others in order to meet my needs. If you are poor, then even if you have needs that I am able to meet, there may be nothing you can offer me that could enable us to enter into a mutually beneficial exchange. Likewise, if you are only slightly poorer than average, but you have needs the satisfaction of which requires the employment of scarce abilities, then you may not be able to enter into a mutually beneficial exchange with anyone able to meet your needs. For it may be that the market sets a high price on the satisfaction of those needs.

This has two implications. One is for the psychology of participants. The pursuit of mutually beneficial exchanges, even if not problematically self-directed, still includes a self-oriented constraint on the services one will be willing, from that motive, to undertake. Call this the direct reciprocity constraint. The other implication is for the principles underlying distribution. Some will have needs to be met but will not meet the direct reciprocity constraint. Such needs will not be met by mutual assistance. If such needs are to be met in a principled way (rather than left to individual discretion and chance), then some alternative principle, and some associated policy or institution, will be required to underwrite this. Call this the restricted application problem.

The direct reciprocity constraint does not just apply in advance of exchange, when one is looking for exchange partners. This constraint also applies once an exchange is underway. An exchange relationship initiated between agents who are in a position to meet the direct reciprocity constraint may end up not meeting the constraint, if one becomes unwell, or becomes committed to caring for dependents, or suffers some other material misfortune. In an important short reply to Sugden and Bruni in this journal, Julie A. Nelson (2009) briefly mentions a case of this latter kind.

Arthur [is] an elderly man with a good pension who needs home visits, and Betty, a woman willing to do these visits for pay with an attitude of ‘friendliness, goodwill, and mutual respect.’ So far, so good. But suppose Arthur suffers a stroke, and thereafter is not able to, himself, express friendliness, monitor the quality of the care he receives, or write cheques. Now Arthur is extremely vulnerable, and is unable to reciprocate for what he receives. At this point, I would argue, important additional social dimensions of economic life, not covered by Bruni and Sugden’s [model of exchange as mutual assistance], come into play. Will the guardians of Arthur’s person and assets continue to adequately provide for his care?

Sugden and Bruni respond as follows (2009):

We can agree with Nelson that authentic caring is sometimes provided in relationships of dependency (as in her variant of our story of Arthur and Betty). When, at the very end of our paper, we offered the tentative thought that reciprocity might ‘go all the way down’, we did not mean to exclude that kind of caring. We intended only to suggest that the authenticity of relationships, even within families and between friends, should usually be understood in terms of reciprocity rather than sacrifice. Sadly, dependency is sometimes unavoidable, but the responses it requires need not be seen as the paradigm of authentic sociality.

Notice that reciprocity is contrasted with (regrettable) dependency in the penultimate sentence. ²¹ Reciprocity is specifically a prospect between those who can meet one another’s material needs, i.e. those who meet the direct reciprocity constraint.

This point of contrast is revealing, since care theorists standardly begin with relationships of dependency and use them as the basis for a more general moral standard. ²² On the approach I am recommending, this identification as humanly vulnerable is the basis for our relationships of caring solidarity. I would say, to mirror Sugden and Bruni, that relationships of dependency and care go (or should go) all the way up, from the relationship between families and neighbours to the institutions of civil society to international politics and commerce.

This is a significant point of difference between the mutual assistance account and the caring solidarity approach. As noted, the caring solidarity approach doesn’t make any assumptions about bilateral usefulness, and certainly not as a condition of service. ²³ A certain approximate material equality may be an implication of a solidarity community but it is not the aim of particular interactions. According to the solidarity approach, each gives what they can and gets what they need. It may be that one’s abilities are modest but scarce, or formidable but commonplace, or modest and commonplace. One’s needs might be considerable or meagre, the means of satisfying them abundant or rare. No matter: the injunction is to do what one can to meet the delegated needs of others. One can live up to this injunction no matter the extent or abundance of one’s abilities or needs.

This is common in friendships. It will often happen in relationships that one party has greater abilities or greater needs than another, for reasons irrelevant to the friendship itself. For instance, having young children means that one’s social opportunities are highly restricted. One’s unencumbered friends tend to have to work around one’s schedule. Likewise, if one party has a significant disability or highly demanding vocation. In well-functioning relationships the parties need not keep a careful accounting of who is doing what for whom. ²⁴ There is enough common knowledge of each one’s disposition to do what they can to meet one another’s needs, so that material equality in ‘effort’ or ‘resources’ is not itself of direct practical relevance. In solidarity collaborations, distribution of benefits and burdens is decided significantly by non-agential circumstance (who happens to need what when and who happens to be able to do what when) rather than by will. ²⁵

4.2 Asymmetric Service and Different kinds of ‘Benefits’

Let me turn now from cases in which parties cannot satisfactorily benefit one another to cases in which they can. We can start with a question about the distributive principle involved in the good case, raised and addressed by Harrison Frye (2023: 25):

One might worry that the mutual benefit model in this sketchy form obscures conflict over the benefits of cooperation. In any given exchange, we might intend some benefit for one another, but there are multiple possible ways of distributing that benefit. For example, you need money, and I need someone to mow my lawn. Mutual benefit is possible, but first we must negotiate. I agree this generates conflict between us, but this sort of conflict by itself is not inimical to communal relations. Even the friendliest of campers are going to negotiate how best to divide up the task of setting up the campsite, clean pots and pans, and so forth.

The word ‘negotiate’ seems to imply that individuals are roughly self-interested in their dispositions to take on this task rather than that (even if their reasons are not self-interested). To continue with the camping example, on this model, one is prepared to do the disjunction of either cleaning pots or setting up camp for the sake of the mutual benefit resulting from these jobs being done. But when it comes to deciding whether to clean pots or set up camp, the individual will endeavour to negotiate with others in order to end up with the task that they most prefer on self-interested grounds. ²⁶

This stance seems problematic to me. On the alternative approach the question of which task to take personal responsibility for is settled by the commitment to living in solidarity with others, and more specifically, to do what one can to meet the needs of others. I know I’m best at setting up camp, even though I’d prefer to be cleaning pots, so I volunteer to set up camp. That would fit better with the ideal of caring solidarity. You know I prefer a certain kind of fish, so you stay out fishing longer until you have caught some. This doesn’t seem at all unrealistic as an ideal for production more broadly. On the contrary, it seems to be a necessary condition of a system of production in which individuals are empowered to fulfil their responsibilities in line with their caring about the needs of one another. ²⁷

At issue here is a contrast between the symmetric distribution of benefits in mutual assistance and the expectations that services will often be asymmetric in caring solidarity.

Exchange is a reciprocal dyad, in which each of the two parties benefits from the interaction. All these dyads are located within one large practice, the market itself. ²⁸ But even in cases in which mutual assistance is possible, and even if there is no room for negotiation (perhaps because both are price-takers, or perhaps because both have transparent and highly specific preferences) – even in such a case, exchange may still not be the most natural or desirable (or efficient) form of the interaction.

In solidarity production, joint activities aimed at mutual benefit will not predominate. In any large economy, they are likely to be the exception. There is no dyadic reciprocity constraint. Most services, in this sense, will be asymmetric, with one party, or rather one group, serving another party, or rather, another group. When you go to the dentist, benefits to them should not be salient. The point of the interaction, the reason you are both there, is for them to fix your teeth. It is not for them to fix your teeth and to pad their holiday fund. Likewise at the baker, the butcher, and the candlestick maker. The team of chefs serves the town hungry. The team of civil engineers serves the local commuters, and so on. All these services can themselves be perceived as contributions to increasingly expansive collaborations instantiating solidarity. In a large productive system, when it comes to interactions between any two participants, only the needs of one of the parties will usually be at issue. In a market system, this will usually be the needs of the one paying the money.

It is economic artifice that these productive services are constrained to be reciprocal transfers – as mutually beneficial in some material sense. We simply assume that when one provides another with a service, the other pays the one money. Plausibly, we assume this because this is how markets work. ²⁹ It fits with an ideal of self-sufficiency, of individuals being responsible for themselves. This, in turn, fits nicely with a morality based on mutual benefit. But if we abstract from the market system, the structure of these interactions is not naturally rendered as mutual benefit. Rather, one person (perhaps in their role in some group) is serving another (perhaps in their role in another group). And ideally, the other person spends their productive time serving others as well.

The ‘economic artifice’ point can be expressed (perhaps laboured a bit) abstractly as follows. Suppose that A has the ability to meet material need n1, B has the ability to meet material need n2 and C has the ability to meet material need n3. And suppose that A needs n3, B needs n1 and C needs n2. Then, if each is disposed to do what they can to meet needs, A will meet B’s need, B will meet C’s need and C will meet A’s need. In this scenario, each one is having their needs met, and meeting the needs of another. But now we can further suppose that B cannot do anything to meet any material needs of A, that C cannot meet any needs of B, and that A cannot meet any material needs of C. On this model, clearly A’s being able to meet B’s need is necessary for their interaction, but B’s being able to meet A’s need is not. If service was premised on mutual assistance, then no-one would serve anyone. An economic model based on exchange requires that each party is able to meet each other’s material needs. In such a model, individuals can be remunerated for service and use that to pay for other services they need, of course. But this seems inessential to the underlying facts about abilities and needs – hence this seems like an artificial structure, and ultimately an artificial set of constraints upon production. ³⁰

Actually, benefit is not an essential part of solidarity interaction at all, either for the ‘agent’ or the ‘patient’. One way to see this point is to notice that A and B can act in solidarity with one another while serving C. We might break this down a little, so that A serves B and B serves C, but A’s serving B is precisely A’s empowering B to serve C. There is still a sense in which A provides a material transfer to B. It is also the case that A facilitates B’s flourishing, inasmuch as B’s flourishing is partly constituted by serving C. We might call either of these a kind of benefit, but they are also other-regarding. In this sense, even the notion of benefit in solidarity ethics involves a kind of breaking down of the self/other dichotomy. In the framework of mutual assistance, there is a breakdown of the self/other dichotomy for motivation but not for benefits. ³¹

Let me clarify this briefly. I’m distinguishing the material transfer from the ‘agent’ to the ‘patient’ from the aretaic significance of solidarity interactions. In the latter context, when you serve someone (in the context of some suitable responsibility), some of your needs are met along with theirs, namely your need to meet needs, your need to uphold your responsibilities, your need to play your part in a caring collaboration, your need for recognition – but also, less abstractly, your need to meet these specific needs of theirs – to teach your students, to bake bread for these people in your community. The same is true of these ‘patients’: their need for recognition, their need to enact a solidarity community, to uphold this solidarity community, are reflected in your manifest care and enacted with their uptake; these points also hold for the various others with whom you are collaborating.

4.3 Charity and Collaboration

In caring solidarity, each is disposed to do what they can to meet the relevant needs of others. But then how is this ideal distinguished from one of charity? ³² It seems one is to be motivated by the needs of others in service, but that would just seem to fall on the ‘other’ side of the old self/other motivational dichotomy.

One crucial distinction between caring solidarity and other-regarding benevolence concerns the role of collaboration. In solidarity, as in mutual assistance, one sees oneself as a team player. One assumes responsibilities as one among others sharing a larger responsibility for coordinating abilities and needs. One also sees those whose needs one is meeting as team players. In solidarity, as in mutual assistance, uptake is crucial. One does not simply impose benefit-enhancements upon the ‘poor’; rather, one shares one’s fate with others. ³³ Solidarity is essentially relational, but charity is often (perhaps always) unilateral. ³⁴

As concerns the self/other dichotomy, here it is significant that living in solidarity is presumed to be among our needs. In fact, a little more strongly, it is presumed to be partly constitutive of our flourishing, assuming we are suitably virtuous. ³⁵ There are multiple dimensions to this. One is the raw moral conjecture that doing what we can to meet the needs of others is part of flourishing. Another concerns the range of freedom in the solidarity ideal. For the collaborative delegation of responsibilities is plausibly responsive not just to abilities and needs, but also to individual interests and commitments. ³⁶ A third dimension is that our individual caring activities are recognized by others and ourselves as worthwhile. This helps us feel good about our work and our community. A fourth is that we can also identify with the larger caring activities of the collaborations that we participate in; this yields stability and existential security. That these different considerations contribute to our flourishing is, of course, controversial. And even if true, the idea that we should structure production in ways that facilitate flourishing, as distinguished from some more liberal conception of empowering discretionary choice, is controversial. But my job here is not exactly to defend a theory of flourishing or a perfectionist ideal in normative political economy as such. My point concerns the relationship between this solidarity ideal and ideal forms of exchange.

4.4 Different kinds of Conditionality

In the solidarity ideal, one is empowered to meet the needs of others. One’s own needs are met, in turn, by others empowered to do so. The same normative expectations apply to the agent and to others and to everyone in the community. The agent normatively expects to be served in their turn. They perceive their service as participation in social production that meets their own needs too.

This ideal involves living aspirationally with others. What we want is for everyone – all who are humanly vulnerable – to live up to the solidarity ideal as best they can. But one’s willingness to serve another, in solidarity, is not conditional upon one’s being served or benefitted by that person in turn. It is no part of the solidarity ideal that we are motivated by a quid pro quo reciprocation, and certainly not in every transaction. This explains, among other things, why the unneedy rich still need to do their bit.

An agent’s willingness to make their contribution is at no point normatively conditioned upon their actually being served. This point is compatible with a number of kinds of conditionality in solidarity. Firstly, as briefly noted above, individuals will have multiple commitments. This is a form of opportunity cost conditionality shared with most moral views – that individuals are at some point conditionally responsive to others with more pressing needs. Then there is a substantive conditionality, not in one’s commitment to living in solidarity, but in one’s commitment to a specific set of collaborative responsibilities. The specific composition of this set of responsibilities is conditional upon its fit within a collaboration that is responsive to a larger set of abilities, needs, and interests. This collaboration is itself responsive to the importance of individuals being able to care about one another, including the importance of their own care being recognized. Finally, there is also a metaethical conditionality: if no-one accepts the solidarity principle, even a little, then it makes no sense to try to live in solidarity. If it is just you, all bets are off. This is an implication of solidarity being a relational ideal; it is most fundamentally a way of living together with others.

Most importantly, the threshold for one’s solidarity participation is not full compliance, but minimal decency and the absence of a universal explicit rejection of the ideal. Also importantly, this is a condition of the society in which one is operating, on whether one should accept a solidaristic morality at all. This is very different from the token-token reciprocal conditionality of exchange.

4.5 Equality as an End versus Equality as Evidence about Will

It is incongruous with the spirit of a caring relationship to focus too closely on accounting for the precise distribution of burdens and benefits in the relationship. This implies that, in solidarity, parties will not insist on benefits to all as the result of any (or every) interaction. It also implies that parties will not have the end of ensuring that the benefits to each party be approximately equal over any particular time horizon.

But it is important not to push this point too far. Feminist philosophers have rightly pointed out that such reluctance can easily serve as a cloak for a gendered division of labour. But there is still a difference here between benefit accounting in a caring relationship and benefit accounting in a joint activity for mutual benefit. In the latter case, joint benefit is the end of the action. The parties are guided precisely by that end in action. But being so directly guided by mutual benefit would be incongruous in a loving relationship, for instance. It would involve a mistake analogous to the ‘paradox of hedonism’. Intuitively, mutual benefit, if it plays any role at all in personal relationships of mutual care, plays a different one. And here’s my suggestion. There is a crucial difference between mutual benefit as part of the end of an action, and as evidence about the kind of relationship that the participants stand in to one another. Feminist philosophers are right to permit us to attend to the distribution of burdens even in very intimate relationships, because that distribution might reveal that one party is not pulling their weight. But it doesn’t follow that the end of their activity is their mutual benefit, exactly. Rather, it is evidence about whether the parties really are doing what they can to meet the needs relevant to a given kind of relationship (close friend, neighbour, teacher, etc.). In fact, the positive disregard for mutual benefit that explains our squeamishness about accounting (even if it is sometimes epistemically necessary) is explained by the fact that we are committed to living in caring solidarity with one another.

This point follows from the fact that solidarity is not a distributive ideal. It is a relational ideal. And, in general, the terms of the specific relationship are what matters most for the recognitional significance of an activity. These can be good or bad. I could provide you with a benefit because you are white, helpless, or my ageing father. Out of love or pity or duty. We have sophisticated semiotic practices to keep track of these further differences. ³⁷ That’s because these further differences can matter enormously to the moral significance of the interaction, and to the kind of relationships, selves, and communities that these interactions enact.

The relational ideal has implications for distributive equality, but only indirectly. This difference can be highlighted by thinking about the very different role of brute luck. As noted, in solidarity interactions, the distribution of benefits and burdens will be decided significantly by circumstance rather than by will. In relationships between individuals with roughly equivalent abilities, needs, and luck, the distribution of ‘burdens’ and ‘benefits’ will be fairly equal. But this fact need not be of any moral significance. It is an unintended consequence of an ideal mode of engagement between parties who happen to have similar abilities and needs and who happen to have similar outcome luck in life.

4.6 The Restricted Application Problem and Domain Differentiation

The domain differentiation that is an inevitable feature of market provision runs against the grain of the ideal of caring solidarity – of all of us working together, and sharing our fates. It is natural to worry that this domain differentiation will lead to a difference in the quality of recognition for service, or social esteem, with dependents and dependency work being respected less than market participants and market-remunerated work. ³⁸

By contrast, the solidarity principle involves principled unity as contrasted with the domain differentiation required by reciprocity-based approaches. This enables continuity across marketizable and non-marketizable services and across interactions between those with market-valued luck and others. Importantly, in the case of solidarity, there is also a kind of continuity in one’s motives inside and outside of the interaction. This can be seen by again reflecting on the difference between isolated relationships and relationships of solidarity. In the latter, the conditionality is built into the role-based expectations themselves. This makes sense. The needs of others you care about include their need to meet the needs of others. There are no significant conceptual differences, on the solidarity approach, between meeting someone’s need, empowering them to meet the needs of another, or working with them to meet the needs of that other together. ³⁹

The ideal would be maximally realized if all of our interactions were to instantiate solidarity with all others. This, of course, includes our relationships in production. After all, we spend most of our lives at work. Our relationships are differentiated by the many different roles we stand in to one another. Only a minority of these roles will be personal relationships like intimate family members or close friends. Most will be vocational of one kind or another, or more broadly economic and political. All can realize solidarity. ⁴⁰

5. Some Economic Implications

Let me summarize and bring a few strands together. Sugden argued that we can perceive exchange as a joint activity aimed at mutual benefit. He also suggests we can see the market system itself, and hence our responsiveness to market prices, as a large-scale joint activity aimed at mutual benefit. On this view, exchange need not be in any way motivated by self-interest. One might never do anything of any moral significance unless and because it is part of a joint activity that promotes mutual benefit. This constitutes an ideal of reciprocity, one that essentially involves seeing oneself as a partner of others in production. One can wholeheartedly value the conjunction, that I serve you and that you serve me. This is all plausible enough. I’m not sure what it is to be an ideal, but certainly this sounds better than egoism.

I have argued that this falls short of an ideal of sociality, on that draws inspiration from the socialist principle: from each according to their abilities to each according to their needs. A society organized in such a way as to honour this principle would meet the needs of those with limited abilities; and it would enjoin the exercise of the abilities of those with limited needs. Furthermore, acceptance of this principle would itself explain why one would continue to serve someone who suffered some financial bad luck after you agreed to help them. For the ideal of mutual assistance, either the same individual would need to rely on a distinct justification (perhaps some separate principle of charity) or else the needs would be left to be hopefully met by charity or the state. Of course, on the other hand, in any plausible ethics, a producer’s commitment to serving someone will not be entirely unconditional. Most people will have multiple commitments. But these are commitments that are themselves worked out in such a way as to accord with the social ideal of living in solidarity. If I don’t give you bread, that’s not because you can’t pay, but because I promised to give it to someone else. No problem there. The main thing is that one doesn’t need to pay much attention to one’s own needs at all in serving others. One’s own needs are rarely part of the point of the exercise.

The primary argument for solidarity concerns the values that it would realize. The most salient is the conjecture that living in solidarity significantly contributes to our flourishing. It would satisfy our needs to care about others, needs for recognition, needs to see oneself as playing a role in a larger worthwhile effort, indeed, in concentrically larger worthwhile efforts. Even if these needs are not universal, they are common, and would become more common in the context of solidarity institutions. I have suggested that this fits with ideas about the importance of care-based identity in social life. Although I don’t have space to argue for this here, I think living in caring solidarity would realize attractive forms of relational equality, ⁴¹ autonomy, ⁴² social freedom, ⁴³ respect ⁴⁴ and an ideal of differentiated unity. ⁴⁵ In short, I think that there is a comprehensive moral approach to social life the centrepiece of which is caring solidarity based on our common human vulnerability.

This ideal is inhibited by market-based economic organization, since the basic structural unit of market-based organization is exchange. This has various implications for the economics of socialism.

Start with market socialisms that aim to restrict markets in non-labour forms of capital, such as Roemer’s (1994) coupon market socialism. This approach has two incommensurable currencies: dollars are used for purchasing commodities including labour; coupons are divided equally amongst adults and used for investing in stocks. This proposal aims to democratize capital management, while leaving a regular market in non-capital goods and labour. Such models would include robust inheritance taxes and perhaps quite stringent rules to ensure effective worker management structures inside firms. Such a model leaves exchange in place as the organizing mechanism for large swathes of economic life. The above contrasts between exchange and solidarity remain in place for these domains – the values at stake in caring solidarity remain unavailable to that extent. ⁴⁶

It is important that solidarity is an ideal that can be realized by degrees. The elements of workplace democracy and economic democracy in Roemer’s model will certainly improve the extent of caring solidarity in production. This is important, given the significance of the efficiency and epistemic arguments for some extent of market provision.

A further important development along these lines is community wealth building. ⁴⁷ Community wealth building (CWB) is a model of sustainable economic development designed to ensure that local economies work for local communities. According to research in this area, CWB ‘seeks to enhance democratic ownership, retain the benefits of local economic activity, and empower place-based economies and workers’ (Lacey-Barnacle et al. 2023). The model was based on the successful Evergreen Cooperative in Cleveland, and a municipally organized effort in Preston. Over 35 local authorities in the UK are practising CWB; the Scottish Government is currently developing legislation to accelerate its use across Scotland. CWB is clearly compatible with market-based exchange and yet clearly also empowers local producers and consumers to recognize that production involves sharing responsibility for meeting one another’s needs in recognizable ways. Like Roemer’s coupon system, this combination improves the extent of solidarity in production while leaving the basic apparatus of market exchange in place.

A more complex market socialist approach is Joseph Carens’ version of market socialism, in which people are encouraged to work for higher gross wages in order to make more significant social contributions, while everyone is taxed 100% and funds distributed in accordance with some desirable pattern. This is a system designed to preserve the epistemic and efficiency advantages of markets while avoiding their distributive implications. Consider the psychology of any particular exchange: I will withhold service from you unless you can pay me. We both know that the wages are going into the social pot, so I’m not really withholding for my own benefit.

One issue concerns esteem. Carens assumes that esteem will be distributed in proportion to the size of one’s gross salary. Carens seems to assume that esteem provides a benefit in a straightforward Sugdenian (et al.) sense. To the extent that individuals are exchanging on the condition that they expect to benefit in this way, this is either an old-fashioned model of self-interested motivation in exchange. Or else, this is a perverse kind of exchange for mutual benefit – and one that feels a bit exploitative.

Alternatively, we could view this esteem as part of a common identification that sustains a desirable form of egalitarian community. Psychologically, esteem would function the way flourishing does in solidarity community: as a positive side-effect of behaviour motivated directly by needs. On this model, the Carensian marketeer takes guidance from market prices about the best way to use their abilities to meet needs in a given context. The market functionally plays the role of a massive collaboration assigning them particular responsibilities. Then individuals are directly motivated to meet those needs (or transfer resources that are means to doing so). I think there are residual problems with this approach concerning recognition: the very opacity of market mechanisms makes it impossible to know why one is being directed towards one service and away from another. ⁴⁸ Consequently, at best, this is an approach that would ape the psychology of altruism, not of mutual assistance. For individuals are following prices blindly, without any structured assurance that they will find the work fulfilling, or any structured support for commitments to specific customers or products. One will be compelled by deference to price to withdraw service from those such as Arthur who become unable to pay. This is another version of the restricted application problem. This kind of invisible responsiveness to a massive system brings with it the forms of alienation that are familiar from criticisms of other impartial benevolence principles. ⁴⁹ So the Carens approach faces a dilemma: either individuals are motivated by old fashioned self-interest, or their motivations are entirely other-regarding; in neither case is the Carens model compatible with perceiving exchange as mutual assistance, nor with caring solidarity.

The epistemic and motivational arguments for the superiority of market economics are standardly used to contrast markets as a decentralized epistemic/motivational system with a command economy based on centralized decision-making. But plenty of epistemic and motivational work already happens with satisfactory efficiency and responsiveness within large private firms and public organizations. What we need are better models of decentralized epistemic and motivational systems based on caring solidarity rather than exchange. The most promising model so far is the Participatory Economics developed by Michael Albert, Robin Hahnel and others (see also Pateman 1970). Generally, though, we are rather epistemically impoverished in the matter of evaluating decentralized alternatives to markets. I hope that the positive project of investigating the relational values at stake in productive relationships will encourage open-mindedness and creativity in developing further economic alternatives.