Justice and Exploitation in Cohen’s Account of Socialism

1 Introduction

A surprising aspect of contemporary political philosophy on distributive justice is the preponderance of liberal theories. Where is the struggle of the labor movement? Where is the fight against exploitation? Where are the socialists? One person who has done his best to change this situation and to give an interpretation of socialism to challenge liberal accounts of justice is G.A. Cohen. He has fought this battle on two fronts. On one side, he has presented some powerful arguments against competing accounts of justice, such as libertarianism and the liberalism of John Rawls.¹ In particular, he has accused Rawls of overlooking the importance of an ethos of justice, and of methodological confusion. On the other side, he has developed a novel account of socialism.² On this account, socialism consist of two principles: the principle of socialist equality of opportunity and the principle of community.

Moreover, Cohen has an answer to the question of where the socialist are, or, at least, where they have been.³ He argues that the reason that socialists, and particularly Marxists, historically found it unnecessary to investigate the nature of distributive justice was that there was a societal overlap between those in need and those unjustly not given their due; the proletariat was doubly disadvantaged. When this was the case, there was no pressing need to examine the relationship between principles of need and anti-exploitation, since regardless of how these principles go together, justice would demand material equality. Due to changes in the economy it is, Cohen claims, no longer the case that this overlap obtains, which means that those in need and those who are exploited are now to be found in different groups. Therefore, socialists have rather recently found themselves having to do political philosophy on the nature of distributive justice and trying to figure out the relationship between principles of need and principles of anti-exploitation. Cohen’s version of socialism is an attempt to take on this theoretical challenge. In this paper, I will investigate Cohen’s account of socialism, while also keeping in mind his criticism of Rawlsian liberalism and the methodological views that it rests on. I will argue that on Cohen’s own theory, his account of socialism comes out as both unjust and exploitative. I will not argue that this implies that socialism per se should be rejected or that the normative implications of Cohen’s theory are misguided, rather I take the upshot to be that the philosophical starting points of Cohen’s project are problematic and in need of revision.

This paper is organized as follows. The next section outlines Cohen’s account of socialism and point to its possibly troubling implication that socialism is not just. The following section is devoted to the structure of Cohen’s arguments against Rawls’s constructivism, and argues that it is indeed an implication of Cohen’s theory that his own brand of socialism is unjust. In the fourth section, I show that socialism also comes out as exploitative, again according to Cohen’s own theory. This is followed by a section, where I investigate a possible solution for Cohen’s troubles with the help of his ethos argument, but find it wanting. In the final section, finally, I will sum up and suggest that Cohen could have avoided these problems of injustice and exploitation by opting for a version of Rawls’s methodology for political philosophy.

2 Cohen’s Socialism

In Why Not Socialism?⁴ Cohen spells out his account of socialism. He invites us to think, not of a social contract in the state of nature, but of a camping trip. A group of people are going out camping together. There is no hierarchy between them, and they have the equipment they need for the trip. They make use of the gear collectively. The goal for this trip is, as usual on camping trips, that each should be able enjoy him- or herself by doing what each most prefer and that they should contribute to the common endeavor as best they can. In other words:

Two principles are realized on the camping trip, an egalitarian principle and a principle of community. The community principle constrains the operation of the egalitarian principle by forbidding certain inequalities that the egalitarian principle permits.⁵

The egalitarian principle is called Socialist Equality of Opportunity (SEO). It is not only “the correct egalitarian principle”,⁶ it is “the egalitarian principle that justice endorses”.⁷ The formulation that justice endorses SEO may make it seem that justice is something else than SEO, but that perception is misleading. For Cohen, justice is a value that is expressed through the principle of FEO. He uses justice and SEO interchangeably (which can be seen in, for instance, in the long quote at the end of this section). Moreover, Cohen is on record as a proponent of the luck egalitarian theory of justice⁸ and SEO is a luck egalitarian principle. For Cohen, SEO is the principle of distributive justice. Justice may sometimes be understood along the lines of an all things considered judgement, but as we shall see in further detail below, when Cohen speaks of justice, he has distributive justice in mind. Cohen gives the following characterization of the implications of SEO.

When socialist equality of opportunity prevails, differences of outcomes reflect nothing but differences of taste and choice, not differences in natural and social capacities and powers. So, for example, under socialist equality of opportunity income differences obtain when they reflect nothing but different individual preferences.⁹

I have emphasized the phrase “nothing but” twice in the quote because this aspect of Cohen’s principle of justice will be important for the argument made below that Cohen’s theory of political morality implies that his kind of socialism is unjust. Before making that argument, however, we must cover some more ground. First of all, SEO is a luck egalitarian principle. As is well known, that kind of principle can lead to some harsh implications, and Cohen is bothered by those. He gives an example of two gamblers who each have $100, and who enjoy socialist equality of opportunity. They flip a coin and the loser pays the winner $50. Given that the choice to gamble was fully voluntary, the resulting inequality is just according to SEO. Since the risks were fully foreseeable, the difference in holdings after the gamble between the two gamblers come about through difference in pure option luck. People are responsible for their option luck, and risky choices that turn out badly lead to justice demanding that people end up in poor circumstances. Severe bad option luck may put people in severe economic circumstances. Cohen notes that even though such inequalities “are not condemned by justice, they are nevertheless repugnant to socialists.”¹⁰ Here we find a first reason for why the principle of justice needs to be constrained by a principle of community.

These kinds of just inequalities can, through aggregation, also create a second problem. Cohen presents a case where he is a well-off man who standardly takes his comfortable car to work, and where another person, who through option luck is less than well-off and therefore forced to take the bus to work every day. Cohen makes about ten times as much as the other person. One day, Cohen’s wife needs to use the car and he has to take the quite uncomfortable bus. This is something that Cohen could complain about to a fellow car-driver, but not to the other person in the example. There is a lack of community between them that is brought about by just economic inequality. The principle of community is needed to avoid this kind of problem. Community requires that “people care about, and, where necessary and possible, care for, one another, and, too, care that they care about one another.”¹¹ This principle, then, directs us to take the needs of others into account. In this it is distinct from justice. The principle of community implies that social circumstances must be such that that people can interact on equal footing. Moreover, it gives expression to an ideal of reciprocity. “Communal reciprocity is the antimarket principle according to which I serve you not because of what I can get in return by doing so but because you need or want my service, and you, for the same reason, serve me.”¹²

The two examples can be used to illustrate how community handles problems for SEO. Say that the losing gambler of Cohen’s example becomes destitute. The principle of community will then imply that the situation must be rectified in some way. The simplest way would, clearly, be to give him some of the money. If the winning gambler cares for the loser, as community directs him to do, then he would find himself obligated to share at least some of his winnings. The outcome of socialism (the combination of justice and community) is different from the outcome demanded by justice. In the bus ride example, the lack of community is brought about by the just differences in economic circumstances. To avoid this situation, the principle of community would, arguably imply that Cohen should be taxed (but see the section “Exploitation and Voluntariness” below for the case of voluntary transfers), and redistribution should take place. With much smaller differences in economic circumstances there could again be community between the passengers on the bus. Again, the outcome after taxes and transfers would be very different from the outcomes mandated by SEO alone. Socialism, for Cohen, is this kind of combination of justice and community.

Above I noted, in explaining SEO, that Cohen uses the phrase “nothing but”, and that I would return to explain why this was emphasized in the quotes. The reason for this is, of course, that if the outcomes demanded by socialism and by justice are different, then the outcomes demanded by socialism are not just. Justice on Cohen’s understanding implies that the distribution should vary with nothing but voluntary choices. He says that “an unequal distribution whose inequality cannot be vindicated by some choice or fault or desert on the part of (some of) the relevant affected agents is unfair, and therefore, pro tanto, unjust.”¹³ In other words, if the pattern of the distribution is not according to something like SEO, then the distribution is unjust. An interesting implication of the “nothing but”-condition is that a distribution that is more equal than what justice demands also comes out as an injustice. Such a distribution would vary with something other than choices, faults, and deserts on the parts of the relevant agents. Now, the original distributions in both the gambling and the bus case are just. The function of the principle of community is exactly to change just but unappealing outcomes into more attractive ones. Socialism demands outcomes that does not solely come about through fully voluntary choice, but that also show community. This leads to distributions that are not endorsed by the principle of justice. Distributions that fail to satisfy the demands of justice are unjust. Therefore, it seems clear that if we share Cohen’s philosophical starting points, we should conclude that socialism leads to unjust outcomes. Cohen, however, seems uncertain.

I believe that certain inequalities that cannot be forbidden in the name of socialist equality of opportunity should nevertheless be forbidden, in the name of community. But is it an injustice to forbid the transactions that generate those inequalities? Do the relevant prohibitions merely define the terms within which justice will operate, or do they sometimes (justifiably?) contradict justice? I do not know the answer to that question. (It would, of course, be a considerable pity if we had to conclude that community and justice were potentially incompatible moral ideals.)¹⁴

Now, our interest here is not whether some transactions should be forbidden, but rather the implications regarding justice of the principles that Cohen endorses as his account of socialism. It seems that regardless of whether the transactions should be forbidden or not, Cohenite socialism leads to injustice since the distributions it demands are incompatible with those sanctioned by the principle of justice. Cohen also expresses uncertainty to whether justice and community contradict each other. Now, there are, at least, two ways in which values can react to combination. One way is straightforward addition. Helpfulness and kindness can serve as examples of this. Doing something helpful in a kind way, does straightforwardly lead to more value being realized. A helpful act does not become less helpful by being done in a kind way, and vice versa. However, community and justice combine in a second way; the combination, arguably, brings forth more value, but this result should be thought of in terms of a net value rather than simple addition. To give an example, it might be best all things considered to combine equality and efficiency, but this combination means a loss of equality.¹⁵ Community and justice often combines in this latter way, which is why socialism can lead to outcomes that are unjust. Outcomes that are sanctioned by socialism, when community has played its constraining role, are unjust according to the principle of justice at the center of socialism. Community and justice imply that different outcomes are justified. This consequence of socialism, that it condemns as unjust outcomes that it sanctions, is what I shall refer to when I say that socialism is unjust.¹⁶

However, if there is a way to combine the principles of justice and community that, in some way can preserve justice while also giving room for community, that conclusion would not follow. What if community was thought to decide what options are available to people (and in this way would come first) and justice is thought of as applying to outcomes within this constrained set, then, perhaps justice and community could avoid coming into conflict. If people could not gamble away their money, then justice would not need to be constrained.¹⁷ I do not think this solution can be made to work in a way that does not have very awkward implications. The gambling case is illustrative of why. If justice condones voluntary gambling and its outcomes, then the implication of restricting options for reasons of community is that just options are taken away and people’s freedom to act within the bounds of justice limited. That doing so could lead to a formerly unjust outcome becoming just, seems very implausible. Why would the restriction of just options lead to unjust outcomes turning just? To return to the example above: if efficiency decides the options, and equality is then brought in, someone like Cohen would not say that resulting outcomes achieve equality. The analogy, here, would be that community first decides the options, and justice, then, comes into play. Now, there seems to be a salient and natural solution to this problem of combining values for Cohen, why not just say that socialism is an account of justice? In the next section, I will show why Cohen cannot make use of this escape.

3 Cohen’s Criticism of Constructivism

In Rescuing Justice and Equality¹⁸ Cohen conducts two distinct but related rescue operations. He aims to save equality from the inequalities allowed by Rawls’s difference principle, and to rescue justice from Rawlsian constructivism.¹⁹ The problem that equality must be rescued from is that if one allows incentives and that people may act self-interestedly, then such an account of equality will allow much too much inequality. The trouble with constructivism is that it mixes up pure justice with other values, which leads to a mischaracterization of what justice is all about. Cohen is a self-professed Platonist and believes that there are fundamental moral principles that we have access to and that are not constructed. In this section, I will present Cohen’s criticism of constructivism and show why it commits him to the position that socialism is unjust. In section 5, we will return to the rescue of equality.

Cohen claims that constructivist accounts of moral deliberation are mistaken. The problem is that the constructivist “identifies fundamental principles of justice with the optimal rules for the regulation of social life.”²⁰ The constructivist account of morality claims that a principle gets its normative qualifications from being the outcome of an ideal decision procedure. In Rawls’s case, this procedure is, of course, the original position, in which the parties are to choose principles of justice based on considerations of facts about human nature, efficiency, publicity and stability. The nature of justice has nothing to do with facts, efficiency, publicity or stability, claims Cohen. There may be good reason to take such things into consideration, but the outcome of deliberations that does so cannot be a principle of justice. Cohen characterizes his problem with constructivism in the sense of a “single disfigurement … from which I seek to rescue constructivism, and that is constructivism’s identification of principles of justice with the optimal set of principles to live by, all things considered.”²¹

To see, in more detail, why this is a disfigurement, when need to turn to Cohen’s theory of value. Fundamental principles, according to Cohen, are neither derived from other principles, nor dependent on any facts. They are the bedrock of normativity. A rule of regulation, however, is the outcome of an all things considered judgment, where the relevant facts and principles have played their respective roles. With these concepts in hand, Cohen claims that Rawls is making a mistake when he presents the principles chosen behind the veil of ignorance as principles of justice, since they are, in fact, rules of regulation. To identify principles of justice with such rules is an error:

My objection to that identification is that, simply, because they are the all-things considered best principles to live by, optimal all-things-considered principles are therefore not necessarily the best principles considered from the point of view of justice alone.²²

Since the difference principle is the outcome of a deliberative process where the parties of the original position come to an all things considered judgment about the principles of justice after taking human nature, efficiency, stability and publicity into account, it is clear that, in Cohen’s theory, the difference principle is a rule of regulation and not a fundamental principle of justice. Each value is what it is and not some combination of other values. For Cohen, real justice is SEO.

Let us now return to the argument of the last section: is socialism unjust? Cohen is very explicit that justice fundamentally should not be thought of as a combination of different values: “fundamental principles of justice are in no way dependent on the character of any facts, or, indeed, and equally important for my purposes, any considerations of value or principle that are not considerations of justice.”²³ Socialism, on Cohen’s view, combines considerations of community and justice. Therefore, socialism is distinct from justice and, moreover, has, by design, different implications than justice. Since socialism is a combination of two principles, it is not a fundamental principle. Since there are only two alternatives in Cohen’s account of values, socialism must be a rule of regulation. If each value is what it is, and socialism is not justice, then the implications of socialism and justice will be different. The outcomes implied by socialism are in direct conflict with the demands of justice, and, consequently, not just. The natural solution of conceiving of socialism as an account of justice is not available to Cohen, given his account of values. Therefore, we can conclude that according to Cohen’s own socialism is unjust. To be clear, the problem with Cohen’s socialism is not that there is trade-off between SEO and community, and it is not that the theory gives the wrong answers. There is nothing wrong with trade-offs and in the two cases we have investigated, the theory seems to give the right answers. Moreover, I have found no fault with the combination of SEO and community per se, or with the account of justice as understood by SEO. The problem is that the theory that Cohen has constructed implies that its own demands are unjust.

Cohen wants to distinguish justice “both from other values and from implementable rules of regulations.”²⁴ We should take him seriously when he says that community and justice are different values. This is why Gilabert’s suggestion that we should think of community as a sufficientarian principle of justice is little of help here.²⁵ It is inconsistent with Cohen’s approach to values, even if it may be a good idea on the terms Gilabert accepts. For Cohen, justice is one value and community another value. This distinguishing approach to values is also why the strategies of Tomlin²⁶ and Lippert-Rasmussen²⁷ for showing that distributive and relational accounts of egalitarianism are compatible is not available to Cohen. Tomlin argues that “all relational egalitarian theories are compatible with some role for distributive egalitarianism”.²⁸ This is, of course, true, but on Cohen’s view the distributive implications of community are just the distributive implications of a value that is not justice. For Cohen, SEO is the principle that justice endorses. The fact that community may have implications for distribution does not show that it is a principle of justice. Lippert-Rasmussen takes Tomlin’s point and develops into the suggestion that relational equality could be reduced to distributive equality. The argument for this is that the claims of theories of relational equality can be redescribed as distributive claims regarding goods like social standing: “any relational egalitarian view is equivalent to a certain form of distributive egalitarianism pertaining to a social good, defined on the basis of the relevant social relation”.²⁹ If community could be reduced to a principle of justice, then socialism may be just after all. However, this move of redescription can hardly be taken to show that relational theories reduce to distributive theories, since if it is true then it is clearly also the case that any distributive egalitarian view is equivalent to a certain form of relational egalitarianism pertaining to a relation, defined on the basis of the relevant social good. Moreover, any Kantian choice can be analyzed as the optimization of a Bayesian utility function, but this does not show that deontological decision-making reduces to consequentialist choosing; it only shows the versatility of the theory of rational choice for purposes of redescription. The reductive approach identifies possible relationships between theories, but it does not provide us with analyses of the content of these values. Cohen’s view is that both SEO and community have distributive implications but that they are distinct values, the reductive approach of Lippert-Rasmussen does not provide us with a reason to revise that position.

Note also that Tomlin and Lippert Rasmussen talk of egalitarianism rather than justice. Now, one could interpret Cohenite socialism as a theory of equality. However, that would only mean that on Cohen’s theory, egalitarianism is the combination of the two distinct values of justice and community. This point can be made even more explicit if we turn to Cohen’s distinction between justice and legitimacy.³⁰ Cohen defends the claim that:

what recommends an outcome that was achieved by just steps from a just starting point is not, in the general case, itself (unqualified) justice, but the different virtue of legitimacy, or, more precisely, the property that no one can legitimately complain about it.³¹

The use of the term “just” in this quote is unfortunate, since in “just starting points” it refers to the justice of a distribution, but in “just steps” it refers to justified ways of changing this distribution. The point of the claim is to distinguish between what justice demands and what can be justified through legitimate steps, and it would have been clearer, but perhaps less illustrative, if Cohen had used “legitimate steps” rather than “just steps” to make his point. This is a very useful distinction, in that it allows us to say about Nozick’s Chamberlain example, that the justice of the (probably) egalitarian distribution D1 is not maintained by the legitimate consensual steps to the inegalitarian distribution D2. Even if the inequality brought about by everyone giving Chamberlain an additional 25 cents at the door is legitimate, it is not just. Using this distinction between justice and legitimacy, we can also say that Cohenite socialism may be an attractive account of egalitarianism, which satisfies the conditions for legitimacy, but that does not show that socialism is just, since justice is not only distinct from community, but also from legitimacy. Cohen is a pluralist concerning values, but he is a monist with regards to the value of justice.

A reply at this point might be to claim that this really is not a problem. If there is a trade-off between values, something needs to be traded off. Moreover, what the principle of community does is that it handles problematic aspects of the principle of justice. If a theory gives the right conclusions, we should not be troubled. We can straightforwardly think of socialism comes as a kind of a rule of regulation. However, from Cohenite starting points, it would seem like a rather severe drawback to have to say that one’s favored conception of justice is unjust. Moreover, this problem of justice gives rise to a further problem concerning exploitation.

4 Exploitation and Cohenite Socialism

The combination of justice and community creates a further and theoretical problem for Cohen’s socialism, namely that it seems to make the theory inherently exploitative. A very general account of exploitation is that to exploit is to take unfair advantage of someone.³² An example, which also could serve as the animating starting point of socialism, would be an employer taking advantage of an employee’s lack of alternatives and paying only a starvation wage. Another example would be if the men on Cohen’s camping trip let the women do all the boring house-keeping work, while they go out on enjoyable fishing trips. When analyzing Cohen’s two examples above, I pointed to some simple solutions for achieving socialist outcomes. In the gambling case, the winner should hand over at least some of his winnings to the losing gambler, and in the bus case Cohen should be taxed so as to achieve the economic circumstances where all passengers can achieve community. However, on the common characterization of exploitation, both these solutions seem to create exploitation. This problem of exploitation emanates from the fact that when we let community override the demands of justice, the resulting distribution will not be just. If justice demands that the losing gambler is carries his own losses, then helping him will make the outcome unjust. This means that the losing gambler enjoys an unjust advantage, while the winning gambler, who has redistributed some of his winnings to the loser, has less than what justice demands. The winning gambler is, in other words, exploited. The suggestion that Cohen should be taxed fares has the same kind of problem. Taxation for redistributive purposes becomes exploitative. When the state taxes Cohen, it takes from him resources that justice says are his and gives them to a person who does not have a justice claim for them. There, then, seems to be a plausible case from standard understanding of exploitation that Cohenite socialism may be exploitive.

However, we have, thus far, only looked at a general characterization of exploitation, but what about Cohen’s own views on the subject? One place to look for his views on exploitation is in “On the Currency of Egalitarian Justice”³³ where he argues for his favored account of the metric of justice – equal access to advantage – by starting from a primary egalitarian impulse: a distribution should not be influenced by brute luck or by exploitation.³⁴ What, then, is exploitation? Cohen says the following. “A person is exploited when unfair advantage is taken of him.”³⁵ We can see the implications of Cohen’s interpretation of this view by turning to an example of his, which is designed to show that welfare is unsuitable as the equalisandum of justice. Assume that equality of welfare is achieved, and that there is one person that cannot be bothered to walk a bit to buy reasonably priced food at the Berwick Street market. Instead he shops at nearby Fortnum’s, a store famous for its exclusive and expensive products. This person will run out of money quite quickly and consequently his level of welfare will be lower than other people’s. In order to restore equality of welfare, we would have to redistribute to this person. Would this be just? Cohen says no, and the reason for this is that it would be exploitative of those who shop responsibly.

It seems to me that, when other people pay for his readily avoidable wastefulness, there is, pro tanto, an exploitative distribution of burden which egalitarians should condemn. Equality of welfare should here be rejected not because of other values but because it is inegalitarian.³⁶

This seems to indicate that exploitation has to do with the resulting distribution. The Fortnum’s shopper is responsible for his lower level of welfare but is still compensated for being worse off in terms of welfare. His shopping habit is an expensive taste, and Cohen’s general solution to this problem is to turn to luck egalitarianism. If the distribution in the Fortnum’s case would have been in accordance with SEO, it would not have been exploitative. This means that we can interpret “unfair advantage” in the characterization of exploitation as not in accordance with justice. This is, hence, is a case the that illustrates the point made above that on Cohen’s view, more equality need not imply more justice, and, indeed, that more equality is compatible with less justice. On this reading both the gambling case and the bus case comes out as straightforwardly as exploitative. In the gambling case, the loser is responsible for his choice to gamble, and in the bus case the original just distribution was not equal due to differences in option luck. According to this view on exploitation, then, Cohenite socialism must be understood as exploitative.

However, there may be another approach to exploitation that could let Cohen off the hook if we turn to his writings on Marxism. In “The Labor Theory of Value and the Concept of Exploitation”³⁷ he outlines a theory of exploitation with a focus on the background conditions for transactions.

Capitalists obtain some of the value of what workers produce because capitalists do and workers do not own means of production: that is why workers accept wage offers which generate profit for capitalists. The crucial question for exploitation is, therefore, whether or not it is fair that capitalists have the bargaining power they do. If it is morally all right that capitalists do and workers do not own means of production, then capitalist profit need not be the fruit of exploitation; and, if the pre-contractual distributive position is morally wrong, then the case for exploitation is made. The question of exploitation therefore resolves itself into the question of the moral status of capitalist private property.³⁸

On this view, which as we have seen is not Cohen’s mature view, exploitation turns on the process which leads up to a distribution, rather than the structure of the resulting distribution. It might be the case that for a full account of exploitation one needs both a problematic process and an unjust end result, but since we have already seen that Cohen’s version of socialism comes out as exploitative in terms of end results, we can focus on the background conditions and the process here.³⁹ Would socialism be exploitative on this approach? There could be one off transfers according to the principles of socialism that are not exploitative. If we start from a just background, and redistribute according to socialism, then this will not come out as exploitative, since the background condition would have, per definition, been just. However, these transactions change the background, and in a society that implements socialism, the background conditions, after the process of redistributive transfers has been carried out, would no longer be just, but instead reflect the combination of SEO and community. Societies that implement socialism come out as coercive over time on this process interpretation of exploitation.

However, Cohen talks about both fair and morally all right backgrounds in the quote above. If we interpreted this in terms of all things considered judgements, we could say that to avoid exploitation background conditions should be justified all things considered, rather than just. This could be one reading of morally all right. If socialism is justified all things considered, then socialism is not exploitative on the process interpretation. There are several problems with this proposal. First, it is far from clear that the all things considered approach is appropriate for making sense of exploitation. Consider a society with an unjust but efficient economy that is justified all things considered. This would presumably be an unequal economy where capitalists would have an advantage over workers; they would have unequal amounts of bargaining-power. If all things considered judgements were what matters for exploitation, this would not amount to an exploitative state of affairs. But on any socialist account of exploitation it is, which points in favor of thinking about exploitation in terms of justice, rather than in terms of all things considered judgements.

It is more difficult to give an example concerning an all things considered judgement based on community and SEO, because it is very easy to fall into a way of thinking that is not allowed on Cohen’s approach, namely that socialism equals justice. Here, however, is an attempt at an example to show that the problem remains even for this combination. Assume that capitalists grow up in capitalist families and thereby acquire expensive tastes, and that workers grow up in working class homes and for this reason end up with frugal tastes.⁴⁰ These expensive tastes could provide us with a pro tanto socialist reason to redistribute from workers to capitalists, which would then create a de facto resource inequality between the groups. This inequality could in turn affect the balance of bargaining power to the detriment of the workers, which would be unjust by the standard of SEO, but that could be justified all things considered by the combination of SEO and community. We might want to say that against such a background, workers would, or at least could, be exploited, but this would be something that we could not say on the all things considered view. This speaks in favor of thinking about exploitation along the lines of justice. So does the fact that the general categorization of exploitation talks about fairness and not morality.

To be clear, the argument put forth in this section is not that it would be wrong of the winning gambler to give some of the winnings to the losing gambler or that the state should not tax and redistribute in the bus case. I do not deny that bad option luck and lack of community could be real problems, and that the kind of solutions that have been proposed seem justifiable. The issue at stake here is philosophical. If a theory of socialism implies that these kinds of solutions are exploitative, then the theory is mistaken. Exploitation is a wrong that has animated the socialist tradition, and one could go as far as to claim that two criteria of adequacy for theories of socialism are that they explain what is wrong with exploitation and that they do not give rise to exploitation. However, Cohen’s theory fails these criteria. Something that is pro tanto wrong, and whose wrongness has animated socialism since its inception is included in the heart of the theory. Cohenite socialism is, according to its own theoretical analysis, exploitative.

5 Exploitation and Voluntariness

Let us now turn to how Cohen aims to rescue equality, and from what. The problem is that Rawls’s difference principle mandates incentives when they are necessary to improve the lot of those worst off. High salaries may provide necessary incentives to get the better off to work hard, which in turn may improve the economic lot of the less well off. However, these incentives may have been made necessary by those who gain directly from them, i.e., the better off, and could have been avoided if they had decided to work as hard without any additional monetary compensation. If the better off had internalized the difference principle, they would not demand higher salaries to do the work they do. The resources used for incentives could have gone directly to the worst off, improving their situation. Cohen claims that since Rawls’s theory is solely focused on the basic structure of society, it has a blind spot for this problem. The Rawlsian idea that justice is the first virtue of institutions is a mistake. The theory of justice should take personal choices into account, and the way Cohen wants to do this is by demanding that each person internalizes an ethos of justice, so that they give expression to this value also in their choices. In this he follows the feminist slogan “the personal is political”.⁴¹ If people truly internalize the difference principle, then they will choose in such a way that the result of their individual choices will be a good as possible for those worst off.

A society that is just within the terms of the difference principle, so we may conclude, requires not simply just coercive rules, but also an ethos justice that informs individual choices. In the absence of such an ethos, inequalities will obtain that are not necessary to enhance the condition of the worst off: the required ethos promotes a distribution more just than what the rules of the economic game by themselves can secure.⁴²

A problem one might have with this proposal, however, is that it seems coercive. Doesn’t this mean that people will have no choice in, for instance, what occupation they should pursue? They should, it seems, take the job that does the most, given their talents and abilities, to improve the situation of the worst off. There is however, Cohen claims, a simple way of solving this alleged problem of coercion. He calls this the ethical solution. It says that freedom is not constrained by moral principles. For instance, we are not coerced by morality, when we abstain from theft or murder for moral reasons. Freedom is not constrained when people act from an ethos of justice. “The ethical solution says that freedom is secured by absence of legal obligation, and equality is secured … through moral and/or quasi-moral commitment.”⁴³ In this way, then, Cohen claims, freedom and equality can co-exist.

On Cohen’s ideal camping trip, there is no state, which means that any redistribution will have to come about voluntarily. People will be motivated through an ethos of socialism to bring about distribution according to community and justice. But when we leave the thought experiment, we need to consider the state. The state will have to implement policies that combine justice and community. Since the state is a monopoly of legitimate force, people will be forced, in at least some senses of the word, to contribute to policies such as the redistributive taxes in the bus case. But as we have seen, such taxes come out as exploitative on Cohen’s theory. The content of the ethos that provides the moral solution may be justified all things considered, but it will then instruct us to bring about outcomes that are not just. This, in turn, implies that following the ethos may bring about exploitation. However, we have also seen how the notion of an ethos has allowed Cohen to claim that the rather strict demands of equality are compatible with voluntariness. If only forced exploitation, such as that brought about by legal rules, is wrong, then there is a way of showing that socialism is not exploitative. One could, then, say that that in order to exploit someone, one must force the person to accept the exploitative outcome. If what seems like an example of exploitation comes about voluntarily, i.e., through the internalization on an ethos, this would not count as an example of exploitation.

This could be a way out for Cohen, but it also creates further problems. To say that voluntary exploitation is not wrong would rob the concept of exploitation of much of its normative force, e.g., we would then have to say that if women acquiesce voluntarily in men’s domination, women are not exploited by men. Consider the division of labor on an actual camping trip. If women do the cooking and cleaning voluntarily, while men do the much more enjoyable fishing, we would like to be able to say that this could be exploitative, even if there is no forcing. Therefore, Cohen should, given his starting points, agree that a criterion for exploitation should not be defined so that only forced gaining of unjust advantages come out as exploitative. If socialism is to be consistent with any kind of feminism, it must recognize voluntary exploitation as wrong.

There is another problem with relying on the notion that only forced exploitation is exploitative. Given the problem of exploitation that we have identified, it seems to follow that only anarchism is consistent with socialism. If the redistribution implied by community is brought about voluntarily through the internalization of an ethos, then socialism does not come out as exploitative, but if it is implemented by the usual means, the state as legitimate monopoly of force, then there is coercive forcing. In this way, the ethos approach would lead us towards anarchism. This solution to the problem of exploitation seems very costly, since it would mean, i.e., that socialism is inconsistent with the welfare state. If Cohenite socialism is incompatible with the state, then it is also incompatible with a redistributive state. Note that this would not be for classical socialist reasons, such the welfare state being an imperfect approximation of the socialist ideal or a hindrance on the road to full socialism, but since it comes out as both justified and unjust according to the principles of socialism. However, the voluntary path is also problematic. If we assume that the state rather is an organization for voluntary co-ordination, then the issue of voluntary exploitation returns. Cohenite socialism seems to be at an impasse. Either it could choose to save socialism from the problem of exploitation by turning to anarchism, which implies that socialism is not consistent with the welfare state and has a strained relationship with feminism, or it must bite the bullet and agree that socialism is exploitative on its own theoretical terms.

6 Concluding Remarks

This paper has made the following claim: the socialist ideal developed by Cohen is both unjust and exploitative on Cohen’s own philosophical theory. In this last section, I want, briefly, to discuss some implications of this result for Cohen’s debate with Rawls and suggest way out of these problems for Cohen. The argumentation here is not meant to be demonstrative, but only suggestive, as it mainly aims to indicate the core reason for the problems I have identified for Cohen’s account of socialism.

In an intuitive egalitarian comparison between Cohen’s and Rawls’s theory, our conclusions point in the direction of preferring justice as fairness over Cohenite socialism. For instance, if one finds it problematic to have to agree to the welfare state being exploitative, one could prefer Rawls’s account of justice to Cohen’s version of socialism. However, this difference between normative theories has its ground in methodological and philosophical considerations. Perhaps surprisingly, these kinds of considerations open a theoretical way out for Cohen, even if it may be costly. It is Cohen’s platonic methodology for arguing about justice that commits him to the position that implies that socialism is both unjust and exploitative. Rawls’s position is that justice can be understood as a combination of several different moral considerations. But justice, claims Cohen, is not a combination of different moral considerations; justice is justice. However, if Cohen would have dropped his opposition to constructivism, a simple solution would have to be open to him. This solution is, of course, to conceive of socialism as based on a constructivist account of justice.

The trouble, as we saw above, is that both in the gambling and the bus case the normatively appealing end results conflict with the demands of justice. If one instead rests content with calling SEO a principle of equality of opportunity, rather than a principle of justice, and conceives, in the manner of constructivism, of the combination of SEO and community as justice, these kinds of problems will not appear. If justice demands that the winning gambler should hand over some of his winnings to the losing gambler, then this is not exploitative. If it is just that Cohen is taxed and the resources redistributed, then there is no account of exploitation that would call this solution exploitative. If Cohen had conceived of the camping trip as a constructivist choice procedure for selecting principles of justice, then the troubling implications of injustice and exploitation could have been avoided.

Cohen would probably have resisted this solution, because on his theory allowing justice to become a combination of other principles turns it into a rule of regulation. However, given what is a stake, it is not obvious that this conception of rules of regulations is worth the costs. Giving it up would let us say the natural thing about socialism, that it is a conception of justice. It would, further, lead to a theory that avoids finding ethical problems where there are none, e.g., finding the welfare state exploitative. Taking the constructivist road at this point dissolves the problems of exploitation we have identified. Of course, there are many more things to consider with regards to the question of choice of methodology, but, at least, these considerations favor the constructivist methodology of Rawls over Cohen’s Platonism.