Luck Egalitarianism, Exploitation, and the Normative Foundations of Socialism

3.1 Why Revisionism Won’t Work

First, let’s consider revisionism – the claim that socialists should simply jettison their commitment to the claim that there is objectionable exploitation in cases like Ant and Grasshopper. The general strategy looks promising at first pass, and indeed this is the approach that Roemer takes, arguing that socialists should abandon the concept of exploitation altogether in favour of a luck egalitarian principle of distributive justice (1994, 65–96; 2017, 274–5). This is, after all, how reflective equilibrium works. Sometimes we have to jettison first-order intuitions when they clash with second-order principles which we have good independent reason to accept.

To see the problem with the revisionist strategy, consider the following case:

Cleanly Generated Capitalism

Suppose we have a society inhabited by ten ants and 1,000 grasshoppers. Suppose that they begin with an equal share of society’s capital stock. Everyone works an equal amount, for equal pay. Suppose then that a few cycles of the seasons come and pass, and through the ants’ superior applications of care and effort, the ants come to accumulate a considerable quantity of capital each, while the grasshoppers fritter away their initial capital stock and end up with no capital holdings. The ants then employ the grasshoppers for long hours and low pay. The grasshoppers make subsistence wages for the rest of their lives, while the ants reap large profits without doing any work themselves.

Intuitively, the situation that obtains in Cleanly Generated Capitalism is one that involves both exploitation and unjust inequality. But, just as in Ant and Grasshopper, it’s hard to see how a luck egalitarian principle of equality could condemn it as such. In line with the thought that drives the problem of exploitation (and which is expressed in the quote from Carling above) our judgement that there is objectionable exploitation in the economic situation that results in Cleanly Generated Capitalism does not seem to depend on historical facts about the genesis of that situation. No matter how a situation like that came about, we have a strong intuition that it represents an objectionably exploitative state of affairs. But the revisionist response we’re envisaging would have to deny this (e.g., Roemer 2017, 272–3).^[13] The revisionist response to the problem of exploitation thus seems to clash with some powerful and important intuitions.

Moreover, we might think that the intuition that such exploitation is unjust even when its etiology is ‘clean’ by luck egalitarian standards is a core socialist insight, one that any putative account of socialism’s philosophical foundations can abandon only at serious cost. There is a long line of thought in the socialist tradition, stretching back at least as far as Chapter XXIV of the first volume of Marx’s Capital, according to which exploitation and injustice can indeed arise through the sorts of clean histories depicted in the Cleanly Generated Capitalism case above.^[14] Indeed, for many, one of the great promised insights of socialist political theory is its potential to explain the lingering sense that the apparent equality that obtains in ‘capitalist acts between consenting adults’ (Nozick 1974, 163) can often mask a more subtle (but pernicious) inequality – the sort of inequality that obtains whenever one person toils to sustain the life of another, who wields power over them and uses that power to extract labour from them.^[15] It is this thought that lies behind the idea, familiar from across the socialist tradition, that capitalism is inherently unjust, rather than simply contingently unjust as a result of the fact that its actual history has been far from clean.^[16] And it is this thought that seems to receive expression in our strong sense that there is exploitation in the Cleanly Generated Capitalism case described above, and that there is something distinctly and objectionably inegalitarian about the situation that emerges from it.

The revisionist response thus seems to carry a very high cost. Though that high cost might still have to be paid if the reflective equilibrium chips fall hard enough in favour of LEP, theorists of socialism’s normative grounds should think long and hard before they conclude that’s the case. If alternative accounts of socialism’s philosophical foundations can be devised which avoid this cost without incurring further theoretical costs in other respects, the problem of exploitation appears to provide us with a strong reason to pursue those alternatives to LEP.

3.2 Why Reductionism Won’t Work

The reductionist method for responding to the problem of exploitation holds that, despite first appearances, the exploitation in cases like Ant and Grasshopper and Cleanly Generated Capitalism can in fact be condemned by the lights of luck egalitarianism. Reductionism may at first seem like a nonstarter as a response to the problem of exploitation. After all, cases such as Ant and Grasshopper and Cleanly Generated Capitalism are, ex hypothesi, instances of inequalities for which the worse-off are entirely responsible. There are, however, two important arguments that it will be worth our while to consider. The first is an argument made by Anca Gheaus (2018) that purports to provide a reductionist method for accommodating certain important kinds of inequalities with luck egalitarian resources, even when those inequalities involve clear cases of responsible choice. The core idea behind Gheaus’s proposal is this. There are some goods which have a considerable influence on how well people’s lives go (and thus on their levels of welfare and advantage) but which are such that they can only be enjoyed by any one individual if they are generally exhibited by society at large. These goods include what Gheaus calls democratic relational goods, or DRGs – such as the goods of living in a society in which there is little social hierarchy, or a significant ethos of solidaristic concern. If DRGs make a significant contribution to the welfare and advantage of particular individuals in a society, Gheaus urges, then there will be a clear luck egalitarian reason for ensuring that these goods are available to those who have not forfeited their right to them through regrettable but responsible choice. In cases where some individuals do make regrettable but responsible choices that land them in situations that threaten DRGs, such as might happen in exploitation cases, we will then have a luck egalitarian rationale for rescuing them, because if we were not to do so we would leave third parties involuntarily worse off than they otherwise would have been, by virtue of depriving them of the welfare and advantage that comes from living in a society structured by DRGs (60–64).

Gheaus’s reductionist approach is an ingenious way of avoiding some of the classic relational egalitarian objections to distributive egalitarianism, like the harshness objection (Anderson 1999; Voigt 2007). But it suffers from a major flaw as an attempted defence of LEP for our purposes. Gheaus’s line manages to ground a luck egalitarian complaint against cases like Ant and Grasshopper. But it provides precisely the wrong sort of ground in the context of debates about the philosophical foundations of socialism. Because according to Gheaus’s line it is only third parties who are the victims of unjust inequality in the two cases. Even if we assume, with Gheaus, that the best way of rectifying this injustice is to protect the worse off from exploitation, this protection will be done for the sake of third parties, who may well be considerably better off. This is both deeply counterintuitive and deeply at odds with the traditional socialist objection to exploitation. That objection is clearly that it is those who are exploited (i.e., the grasshoppers in Cleanly Generated Capitalism) who are the victims of unjust inequality, not (potentially privileged) third parties. A Gheaus-style reductionist response to the problem of exploitation thus secures the claim that there is unjust inequality in cases like Cleanly Generated Capitalism only at the significant cost of (a) an implausible account of the nature of that unjust inequality, and (b) the betrayal of the spirit of the core intuition that drives the problem of exploitation in the first place.

The second reductionist strategy employs Serena Olsaretti’s work on luck egalitarianism and the question of stakes (2009, 2013). For our purposes, Olsaretti’s central point is this: luck egalitarianism might be correct to attempt to accommodate the idea of responsibility by insisting on the significance of the choice/circumstance distinction. But Cohenite luck egalitarianism is married to an implausible view of what the stakes of people’s choices should be. Though luck egalitarians may be right to suppose that equality requires that people face the consequences of their responsible choices, there is no uncontroversial way of deciding the appropriate size and character of those consequences. Thus, one might think that equality does require that somebody in the pit case should be worse off than others if they end up trapped in a pit through their own responsible choices. But one might think that equality condemns a situation in which they end up so badly off that they are forced to sign a sweatshop contract. In other words, one might think that equality is consistent with people being worse off than others when their situation is the result of responsible choice without thinking that it is consistent with them being that much worse off. The stakes, one might think, shouldn’t be that high. The implication for our debate is that one might think that LEP can avoid the problem of exploitation by making use of a more plausible theory of stakes – of the consequences that it is reasonable to require others to face as a result of their responsible choices. If the most plausible version of luck egalitarianism holds that being forced to sign a sweatshop contract is too high a cost to impose on the choice to, for example, spend the summer singing rather than working, then perhaps luck egalitarianism can condemn the exploitative offers in cases like Cleanly Generated Capitalism as contrary to the demands of equality. Those cases would transgress equality by imposing too harsh a penalty on certain kinds of choices.

A modified LEP with a different theory of stakes is a serious contender for a satisfactory form of reductionism that could rescue the LEP model from the problem of exploitation. Nevertheless (and although a full exploration of this possibility would take us too far afield here) I think there are some good preliminary reasons to doubt that this strategy could succeed either. The main obstacle to a stakes-based reductionism is that the resultant view seems to collapse into an alternative to, rather than a version of, luck egalitarianism.^[17] The first problem is just that the view will now contain an alternative component – the attached theory of stakes – that may well answer to values other than equality. For example, Olsaretti’s preferred solution to the problem of stakes is to ground a theory of stakes in something like Cohen’s principle of community. Penalties should be attached to choices in such a way as to maintain community as best as possible over time (2013, 62). On such a view, luck egalitarianism is no longer an attempt to articulate only what equality requires; it is, at least, considerably shaped by the requirements of community. This represents a serious departure from the original luck egalitarian program pursued by the likes of Cohen and Roemer. Indeed, as a response to the problem of exploitation, it seems better described as appealing to the plus than reductionism, and so it seems likely to run into the criticisms of appealing to the plus raised in the next section.

The second problem is that the additional component will most likely introduce new requirements that will pull even further away from the original impulses of luck egalitarianism. On a communitarian theory of stakes, for example, the appeal to Cohenite community seems to undermine placing the sort of emphasis on the choice/circumstance distinction that characterises the central luck egalitarian impulse, since, as Cohen argues, his conception of community condemns even those inequalities of condition which arise from regrettable choice or option luck. Or consider a plausible alternative to Olsaretti’s preferred approach to stakes: suppose we were to develop a theory of stakes that is designed to maintain not communal relations over time, but egalitarian relations, attaching penalties to choices that will minimise the emergence of inegalitarian social relations by preventing power inequalities, discouraging hierarchies of esteem and so on. Such an approach may well provide a plausible theory of stakes. But it’s not clear why we would want to describe this theory as a luck egalitarian account of equality, as opposed to a relational egalitarian one. Moreover, again, if we’ve admitted that relational egalitarian considerations are in the driving seat for a theory of stakes, shouldn’t our sensitivity to those considerations also give us reasons to care about relational inequalities quite apart from considerations of the choice/circumstance distinction?

Olsaretti’s revisions to the classical luck egalitarian program thus seems likely to yield a very different theory, with far fewer counterintuitive implications. But precisely because its spirit and content would be so different from classical luck egalitarianism, it’s hard to see why we should describe such an account as a version of luck egalitarianism at all, rather than simply as an altogether alternative account of what equality requires (or indeed an account of something else entirely.) And it’s therefore hard to see how it could be used to defend a distinctly luck egalitarian approach to socialism. Indeed, the two reductionist strategies just overviewed suggest that the reductionist approach to the problem of exploitation faces the following dilemma. Either the novel account of luck egalitarianism’s requirements cleaves too closely to the original, and so retains its counterintuitive implications, or it avoids those counterintuitive implications by revising the original so heavily as to undermine any claim to continuity with it. As an attempt to rescue the LEP model of socialism, appealing to democratic relational goods falls on the first horn of that dilemma. Stakes-based reductionism falls on the second. The prospects of a successful reductionist response thus look slender.

3.3 Why Appealing to the Plus Won’t Work

In the previous two sections we reviewed the prospects of revisionist and reductionist methods of resolving the problem of exploitation and found them unlikely to yield satisfactory responses. This leaves our third and final method: appealing to the plus. Such a response could start by pointing out that in cases where exploitation is the result of luck egalitarian inequality (and, we might add, this might be thought to cover the vast majority of important real-world cases of exploitation), there is a straightforward luck egalitarian complaint of unjust inequality against the exploitative situation. But even in contrived cases like Ant and Grasshopper and Cleanly Generated Capitalism, where we may have no grounds for a complaint on grounds of equality, we may still have a complaint on grounds of community or solidarity.

Appealing to the plus seems an attractive way of dealing with counterexamples like Ant and Grasshopper and Cleanly Generated Capitalism. Though those cases don’t seem to involve any luck egalitarian inequality, they do seem to involve clear breaches of solidarity and community. With regard to Cohen’s principle of community, the exploitative offers that are extended clearly fail to instantiate communal caring and reciprocity. With regard to Roemer’s principle of solidarity, the offers likewise do not seem to be instances of solidaristic thinking. The ants are clearly seeking to maximise their own advantage in their interactions with others, rather than seeking to play their part in a coordinated response to a collective problem.

With these ideas to hand, the defender of either Cohenite or Roemerian LEP might be able to respond to the problem of exploitation as follows. LEP may not be able to object to all cases of exploitation on grounds of equality, but (a) it can provide an egalitarian objection to any exploitation that does involve luck egalitarian inequality, and (b) even in those (rare) cases where luck egalitarian inequality is not involved, LEP can still provide a distinctly socialist objection to exploitation on grounds of community or solidarity, instead of equality. LEP can thus ground a principled socialist objection to exploitation, per se. And if so, why does it matter that it cannot ground that exploitation on the value of equality exclusively? Though equality is certainly a core value in the socialist tradition, it is by no means the only such value. It should therefore be no embarrassment to a putative account of socialism’s philosophical foundations if it doesn’t manage to ground every first-order socialist commitment on the value of equality alone.

However, there are two important problems that face attempts to use appealing to the plus as a way to defend LEP from the problem of exploitation. First, if the proposed LEP account of socialist opposition to exploitation in cases such as Cleanly Generated Capitalism is correct, then in those cases such opposition is grounded not in equality but in community or solidarity. But given that Cohen and Roemer both make equality, but not solidarity or community, a requirement of justice, it follows that there can be no socialist opposition to those cases on grounds of justice. And this might be thought to generate a problem for the following reason. There is a long and widely held view in political philosophy according to which justice bears a special priority over other values in the evaluation of social institutions. In Rawls’s words, there is a strong ‘intuitive conviction of the primacy of justice’.

Justice is the first virtue of social institutions, as truth is of systems of thought. A theory however elegant and economical must be rejected or revised if it is untrue; likewise laws and institutions no matter how efficient and well-arranged must be reformed or abolished if they are unjust … The only thing that permits us to acquiesce in an erroneous theory is the lack of a better one; analogously, an injustice is tolerable only when it is necessary to avoid an even greater injustice. (1999, 3)

Now, a more sophisticated theorisation of this rough articulation of our intuition of the primacy of justice is no doubt in order to work out the precise parameters and implications of such a view. But the rough articulation suffices for our current purposes. Because if justice does hold the sort of primacy indicated by Rawls, shifting the grounds of exploitation charges from equality (and thus justice) to solidarity or community will have important implications for the status and quality of those charges. Though they may still carry a good deal of moral weight, they will not carry the same sort of significance as justice claims and will have to cede to such claims in cases of conflict.

The force of this problem should not be overstated. Since LEP will still be able to ground all cases of exploitation which do involve luck egalitarian inequality in complaints of justice, it is only those fringe cases of cleanly generated exploitation that LEP will have to criticise on grounds other than justice. Moreover, one might think that the primacy of justice has less purchase in the socialist tradition than the liberal anyway; Cohen (2009) strongly suggests that justice will not always take priority over other values in his brief discussion of value conflict, and it might be argued that the traditional socialist emphasis on the value of solidarity and community contains the seeds of the rejection of the primacy of justice for similar reasons (37).^[18] However, the problem is nevertheless significant. For if the arguments of Section 3.1 are correct, those engaged in the task of producing normative foundations for socialism should be very reluctant to jettison the idea that cases like Cleanly Generated Capitalism involve moral wrongs with the same kind of moral weightiness as cases of exploitation which involve straightforward inequality of opportunity along luck egalitarian lines. Given the intuitive power of the idea of the primacy of justice, the socialist theorist who wishes to ground the wrong involved in Cleanly Generated Capitalism on values other than justice must either provide us with a reason to reject justice’s primacy or abandon the important intuition that exploitation cases like Cleanly Generated Capitalism retain their moral character across different etiologies.

This brings us to the second criticism of appealing to the plus as a method for responding to the problem of exploitation. Irrespective of the primacy of justice, appealing to the plus saddles the defender of LEP with an account that provides a poor fit for important socialist judgements about cases like Ant and Grasshopper and Cleanly Generated Capitalism. In particular, appealing to the plus seems to clash with two important judgements about such cases: an intuition regarding their relative moral parity when compared with cases of exploitation which involve luck egalitarian inequality; and an intuition regarding the distinctly inegalitarian nature of the wrong in such cases.

With regard to the first intuition, appealing to the plus involves arguing that amongst cases of objectionable exploitation there is an important moral fault line, dividing cases which involve luck egalitarian inequality, and can therefore be condemned on grounds of equality, and cases which do not involve any such inequality, like Cleanly Generated Capitalism. But this clashes directly with the intuition that our moral judgement of exploitation cases does not vary according to etiology, or at least not along the lines of the choice/circumstance distinction that is so central to luck egalitarianism. No matter how the grasshoppers end up propertyless, or no matter how the entrapped person finds themselves at the bottom of the pit in Carling’s pit case, we should have a similar objection to the exploitation that occurs in each case. (This is not to say that certain, particularly objectionable, etiological differences might not still make an important difference in these cases. See Section 4 for further discussion.) That objection does not seem to depend on prior facts about the history of the cases, but rather the bare facts of how the individuals in each case interact and relate to one another – relational facts which are insensitive to etiological variation along the choice/circumstance divide.

With regard to the second intuition, appealing to the plus also faces the significant problem that it involves a flat denial of the claim that there is a distinctly egalitarian critique of the exploitation that we find in cases like Cleanly Generated Capitalism. But such cases, as we saw in Section 3.1, don’t just seem to be wrong; they seem to involve a wrong with a distinctly inegalitarian flavour. The relationship envisaged between the ants and the grasshoppers is one in which the latter are effectively rendered the servants of the former.^[19] To classify such wrongs as failures of community or solidarity – failures to care adequately for one another or to conceive of one’s situation in adequately collective terms, rather than failures to relate to one another as equals – is to misdescribe the moral quality of the situation. The evils of Cleanly Generated Capitalism do seem to exhibit the distinctly inegalitarian ills of hierarchy, domination, and subordination. The grasshoppers do not seem to only have a claim that they have not been extended adequate community or solidarity by the ants (though they certainly have that too). They also seem to have a claim against being treated as the ants’ subordinates and inferiors. If these intuitions about the moral quality of the wrongs involved in cases like Cleanly Generated Capitalism are correct, then appealing to the plus seems ultimately unattractive as a response to the problem of exploitation. It purchases that response only at the considerable expense of an account that seems to fundamentally miscomprehend the moral nature of the core phenomenon.

This final point is also helpful for showing why an alternative kind of ‘appeal to the plus’ is similarly unpromising. According to this thought, the sorts of relations envisaged by ideals of solidarity and community are not substantially different from those envisaged by certain relational ideals of equality (Tomlin 2015, 153–4; Roemer 2017, 310). As such, we might be tempted to redescribe the ‘plus’ component of LEP as in fact an egalitarian requirement – a requirement of relational equality, sitting next to the distributive equality mandated by luck egalitarianism. Particularly if we amend the standard formulation of LEP to hold that the plus component is also a requirement of justice, we might think that LEP can appeal to the plus without running into the objections presented in this section – Cleanly Generated Capitalism will be condemned as unjust just as much as unclean cases, and though there will be a difference in the specifics of the normative diagnosis of cases that fall on different sides of the choice/circumstance line, that difference will not track a distinction between egalitarian and inegalitarian objections.

My first response to this suggestion is to point out that the resulting picture has come some distance from the LEP models of Cohen and Roemer, so much so that it seems a stretch to describe them as instances of the same model (employing, as they do, fundamentally different understandings of justice, equality, community, and solidarity). But setting that aside, it’s also worth noting that the attempt to subsume an ideal of relational equality within an ideal of communal or solidaristic relating is likely to meet some serious problems, because the sort of communal concern and care associated with solidarity and community seems to be importantly distinct from the equal respect and equal status associated with relations of equality. The most obvious way to put the worry is that, at least on the face of it, it doesn’t seem difficult to think of social relations that exhibit considerable degrees of Cohenite communal caring or a Roemerian cooperative ethos without the parties to the relevant relations relating to one another on equal terms, and vice versa. Think, for example, of a caring relationship between a husband and wife in a patriarchal society, or, for the inverse, two rivalrous boxers eyeing each other before a bout, brimming with respect for an equal adversary, but intent on thwarting each other’s ends.

The philosophical waters start to run a little deep here, but the basic thought can be illustrated by indulging a spatial metaphor: while failures of relational equality involve vertical relations of higher ups to lower downs, failures of solidarity or community operate on a horizontal dimension, of closeness and distance. In Niko Kolodny’s words, the point of ideals like solidarity and community is ‘to reject separation, rather than superiority’ (2023, 265; see too Lippert-Rasmussen 2018, 234). Though much would depend on details that are beyond the scope of this essay, these considerations suggest that equality and solidarity or community do not naturally pull together as relational ideals; while they might complement each other, and while each might be vitally important to socialists, they refer to fundamentally different kinds of relation, and fundamentally different values. This gives us even more reason to suppose that if a modified version of LEP could be concocted on which the plus component requires egalitarian social relations as a matter of justice, and thus condemns cleanly generated exploitation as unjust and inegalitarian, it would be a theory that lies some distance from the original LEP model of Cohen and Roemer, and would most likely be better thought of as an alternative to it than as an instance of it.