Ethics and Capitalism
September 28, 2021
Here I want to outline an argument for the claim that ethical action is impossible under capitalism. This is the kind of claim that is liable to get one labelled as a hopeless cynic—unfortunately for me I happen to believe it, so the aim here is really to clarify the claim and its rationale in a manner which opens as many doors as it closes. There are two main components. One is the argument that act utilitarianism is self-defeating, because a community of act utilitarians will be unable to solve moral coordination problems (a criticism which has historically served as a motivation for rule utilitarianism). The other is a claim to the effect that when our ethical life is expressed through market activities, this makes us de facto act utilitarians, regardless of our de jure stance on normative ethics. Roughly, then, the argument is that the mediation of ethical life by the market systematically undermines our capacity for moral coordination. The final premise is that morality is essentially a matter of coordination—that without moral coordination there is no morality at all.
First I should clarify what the claim isn’t. When I say ‘action under capitalism,’ I do not mean to refer to any action which could conceivably be taken while capitalism is the dominant economic system, which in our context would mean all action. This would be tantamount to claiming that ethical action as such is impossible now, full stop. I do not believe this. Rather, I want to defend the weaker claim that any action ‘functionally subsumed’ under the capitalist mode of production is divested of its ethical content, in a sense soon to be clarified. By this I just mean any act subsumed under commodity production, consumption, or exchange. An obvious example is ethical consumerism. But this also includes ethical production—for instance the decisions one makes about one’s job, insofar as this involves contractually obliging one’s labour to the production of commodities.
The boundaries get very murky here. What actually counts as commodity production these days? What if you work in a university where your time is split between government funded medical research on the one hand, and preparing students for the labour market on the other? What if the start-up you work for spends half of its time doing exciting world-changing development stuff, and the rest doing corporate design work to fund it? What if you’re a software developer and you make contributions to open source projects, but then use them as CV points to find yourself paid gigs in commercial tech? Does that mean the open source contributions, which in themselves do not constitute market activity, have been subsumed into commodity production via the role they now play in your professional branding? What if you leverage inherited wealth to set up a charity, make a big noise about it being a cooperative not-for-profit, then set up a separate business in your own name where you cash in the social capital gained from the first venture to secure lucrative speaking gigs? What if you quit a high-profile job for sincere ethical reasons, then use the experience to write a bestseller about the precedence of morals over profit? If this is a legit example of functional subsumption, wouldn’t that imply that non-market activities can be retroactively subsumed into the market by future actions? If so, then what about when IBM adopted the Debian open source operating system, effectively capitalising on the free labour provided by anonymous developers in the past? If not getting paid doesn’t mean you’re not producing commodities (or future commodities), then where does an activity like posting on social media stand in the landscape of generalised commodity production?
Clearly there’s a whole host of thorny issues here, and many details I shall pass over. I will however note two things. The first is that these are not simply questions of classification or propriety—they concern real incentive structures with empirical consequences. It makes a difference whether your open source contributions are aimed at making robust, genuinely useful software or boosting your CV: if it’s the latter you’ll be pushing experimental features for flashy projects that may ultimately go nowhere, while the former is more likely to involve vast amounts of unglamorous work ironing out bugs in the file management system. The second thing to note is that for many, questions like these represent exactly the kind of trade-offs that make up the fabric of contemporary ethical life. This is just to reiterate the (perhaps obvious) point that moral decision making is largely and increasingly a matter of deciding how to trade one’s labour and its spoils on the market, a statement intended less to pass judgement than to state a fact of life.
Now to the core argument. Act utilitarianism refers to the most familiar form of utilitarianism, contending that the right action to take in a given scenario is that which maximises expected utility, and that this should be decided on a case by case basis. This can be contrasted with rule utilitarianism, which contends (like the deontologist) that we should act in accordance with moral rules, and (unlike the deontologist) that the rules we should act in accordance with are those which would maximise overall utility if adhered to by all. What’s important here is that for the act utilitarian there is no such thing as an unconditional moral value—the act utilitarian can never say, for instance, that killing is wrong full stop, because we can always imagine a scenario in which killing someone increases expected utility (for instance by saving the lives of five others). The rule utilitarian, on the other hand, can say simply that killing is wrong full stop (even when it would save more lives) because universal and unconditional adherence to a rule against killing is the best way to maximise overall utility in practice, even though we can construct edge cases where it does not.
The tension between act utilitarianism and coordination was first pointed out in a 1978 paper by Allan Gibbard. Many detailed responses and revisions have followed, but here I’ll just lay out the thrust in broad strokes. Generally, the idea is that act utilitarians can’t coordinate because they can’t assume unconditional values in each other. If you have a population of act utilitarians who understand each other as such, then no-one can be sure that others will stick to their moral commitments, because as act utilitarians these commitments are always revisable on the basis of new information, and never serve as norms which constrain action unconditionally. In this sense they are not really ‘commitments’ at all, in the common sense of the term—they are more like provisional advertisements of likely future behaviour, but without actually exercising any binding authority over them. But, for standard prisoner’s dilemma type reasons, this is insufficient to secure cooperation in many situations, and will yield races-to-the-bottom with morally suboptimal outcomes (by their own lights). Hence, act utilitarianism is self-defeating in practice.
That’s more of an outline than a detailed argument, and clearly more needs to be said, but for present purposes I’ll assume it is on to something. Immediately there’s two things to note. The first is that this whole argument only applies to cases in which moral action depends on group coordination, so one obvious response is that most situations which demand moral decision making are just not like this. And indeed this is one of the main ways that act utilitarians have pushed back against Gibbard-style arguments—I’ll return to this point later in the post, where its relevance to the present concerns will be clearer. The other thing to observe is that the failure of coordination does not originate in everyone being act utilitarians—it originates in their perception that everyone else is an act utilitarian. In the given example this perception is stipulated as true, but coordination would equally fail if everyone were to have the false impression that others are act utilitarian. Even if they were in fact rule utilitarians, they would not be able to establish as common knowledge any shared rules around which to coordinate. This will be important in the next stage of the argument.
It’s probably clear where this is going: what I want to say is that when an act is subsumed under the capitalist mode of production it can no longer communicate an unconditional value. This is because to be so subsumed is to be assimilated into general commodity exchange, and what is distinctive about commodity exchange is that it occurs under the sign of universal equivalence. Or in more colloquial terms, it is to put a price on moral values, to hold moral values commensurable with each other in practice by trading them against a single scale—i.e. it is to become a de facto act utilitarian. The sense in which ethical action is impossible under capitalism, then, is not so much a claim about particular acts. All else being equal, it is no doubt more ethical to buy the one thing rather than the other. But all else is never equal—the contention is that the more our ethical life is conducted through market activity the less capable we become of coordinating our ethical decisions, precisely because it means we can no longer communicate unconditional values to one another. Ethical life becomes intrinsically atomised, or privatised, each decision wrapped in an implicit ceteris paribus clause–a million individual moral wins are bought at the price of undermining the possibility of establishing a universally binding moral consensus. Since this universally binding quality is essential to moral life, this is tantamount to the deterioration of ethical subjectivity as such, or a failure to synthesise the universal subject (manifesting practically in the collective inability to avoid moral prisoner’s dilemmas and Malthusian spirals).
It’s probably worth saying a bit more about this de facto/de jure distinction. Let’s say you are a de jure deontologist or rule utilitarian—i.e. this is your actual stance on normative ethics. Let us add to this that you are an ethical vegan, and you take minimising harm to all sentient beings to be an unconditional moral rule. Naturally, these beliefs will be expressed through your market activity, by the products you buy and perhaps in the career you seek or the content you produce. The point here is that none of these activities can communicate the unconditional universality of the rule as you understand and intend it—when you spend £8 on the seitan and seaweed burger rather than £4 on the fish-finger sandwich all you are communicating is that you value that ethical gain at £4 at least. It does not communicate that you wouldn’t have bought the fish-finger sandwich if it had cost £1 (say). The only way to communicate a unconditional value would be to pay an infinite price for it, which is impossible, or perhaps to trade a priceless object for it, which would fall outside of commodity exchange by definition. Even if we do indeed hold unconditional commitments, our market activities can only represent them as commitments negotiable against other commitments on a single scale of value. The market ‘reveals’ us to one another as act utilitarians, even when we are not act utilitarians. By the argument made earlier, this is sufficient to undermine coordination.
At the centre of this line of thought is the idea of the market and money system as a medium of communication. As fans of Austrian school economics like to point out, our market activities of buying and selling (including selling our labour) express our value hierarchies, and it is in virtue of this information transfer that markets can be understood as computational systems. But what is often left unconsidered is that by expressing our value hierarchies within the money system we restrict ourselves to expressing those values as if they were strictly commensurable, existing on a single, continuous linear scale. Whatever is gained in the frictionless information transfer of the market is bought at the price of excluding incommensurable values, such as hard moral rules—but these are critical for certain kinds of coordination. In his paper On Technical Mediation, Bruno Latour argues that mediation can be understood as a form of translation—specifically, as the translation of ends (Latour, 1994). What we find in the mediation of ethical communication through market activity is exactly such a translation: the translation of a language of value with one set of expressive capacities (including the capacity to express unconditional or incommensurable values) into a different grammar with different capacities (one which can only express values as strictly commensurable with one another). The site in which translation occurs is not that of the subject, not in our intentions; it lies rather in the objective reading conventions materialised by the formal structure of the medium itself.
One objection to this line of thought might be that it assigns far too much weight to the market as a means of communication. Even if we accepted that moral communication mediated by the market is unable to secure moral coordination, couldn’t we just reply that (clearly) we use all sorts of different communications channels for moral deliberation? We can just talk to each other, surely? The answer, I think, is yes: of course different communication ‘spaces’—different expressive media with different grammars—including good old fashioned peer-to-peer private speech, don’t suffer from the problem I’ve outlined here. But the question is not could we do this, but do we do this.
Earlier I reeled off a big list of fictional-not-fictional scenarios in which the line between market and extra-market activity is blurred. Let’s take the example of the person who quits their job for moral reasons, then writes a book about their experiences which ends up selling well. We can imagine a totally sincere pathway through this story. The choice to quit is a heartfelt response to perceived complicity in unethical practices. Staying silent then seems like further complicity, since if no-one speaks out then everything will carry on as before. They talk to people about it privately, but this doesn’t seem to make any difference to anything: making a public statement seems the only way to genuinely raise awareness of the problems in this industry. The message can achieve nothing if it is not heard. Turns out lots of people feel similarly and resonate with its message—this is exactly why it sells so well. Nowhere is there any self-serving intention or insincerity, but nevertheless the moral conviction is ultimately expressed through a (successful) act of commodity production.
What has actually been communicated here? Ultimately: a transaction between one kind of capital (the original high-profile job) to social capital (the clout one acquires in the public eye for having taken a stand) back to regular capital (book sales). At every point of this series, each store of value is exchanged for another on the basis of some shady but ultimately unified scale. The expressive problem with markets dealing in material goods and money seems extendable to markets dealing in social capital—in fact, this is exactly (it seems to me) the difference between modern day clout—which refers to something fungible—and traditional prestige, which only has meaning in the context of a concrete community and doesn’t make sense as a tradeable ‘value’ on the open market.
I won’t dwell on this point, because the post is already long. The point is really only that to the extent that our moral activity is expressed through our market activity it is essentially hamstrung with regard to moral coordination. But if we are honest, I think we shall have to admit that an enormous (and always increasing) amount of our communicative activity—posting on social media, even private communications with each other in spaces structured by professional norms, etc—is subsumed by commodity exchange, in one way or another, which implies that it is constrained by an expressive grammar which places all values on a unified linear scale.
So that’s the bare bones of it. There’s a couple of things brushed over here, not least the premise that moral coordination is essential to morality as such. I must admit that here I’m smuggling in some significant (perhaps Kantian?) assumptions—coordination is essential, I believe, because it is what is required for the universalisability of the moral principle to become a practical reality, not simply a regulative ideal. It is through moral coordination that a collective subject is realised as a concrete reality. Without it moral talk becomes empty, because the universal ‘we’ implicit in the moral principle indexes no concrete community—it is precisely through its disassembly of the concrete universal subject that the market undercuts moral life, in effect reducing it to its epiphenomena. But these are just sketchy commments—this part of the argument I leave as a placeholder, to be fleshed out in the future.
As a final point I’d just like to defend my original claim that this argument is not cynical. By a cynical argument surely we mean an argument that critiques without providing alternatives, an analysis that leads to an apathetic mire with no recommendation for action. So if my argument is not cynical it should provide some recommendations as to what to do, of how ethical action could be possible now even if ethical action is impossible under capitalism. And actually, the recommendation is quite straightforward. The problem with mediating ethical activity through market activity is that it means we can only express our value hierarchies on a continuous linear scale, but moral coordination requires the communication of discontinuities in our value hierarchies: unconditional values and deontic redlines. The required strategy, then, would be to find a way of recovering the possibility of expressing such discontinuities. As I have already said, this can’t be achieved by buying things. Nor can it be achieved by selling things—however, it can be achieved by refusing to sell something. To refuse to sell something at any price is precisely to communicate a value that cannot be located on the linear scale of the market. It is to make it priceless. Since the foremost thing any of us has to sell is our labour, it is in this domain that this form of resistance is best exercised. This is why I believe the Italian Autonomists were extremely on point in their adoption of the refusal of work as a political tactic. If the contemporary situation is more complicated, it is only because the exchange of social capital often undermines the politics of refusal in subtle ways, precisely by distorting the medium through which values are communicated. By my argument, to quit a job on ethical point is an ethical action—it successfully expresses an unconditional value—but not if the action is traded in for social capital, then or later. What is required, then, is a two stage refusal: first the refusal of the money capital for one’s labour, then the refusal of the social capital one can expect in return for the primary refusal.
References
- Latour, B. (1994). On Technical Mediation—Philosophy, Sociology, Genealogy. Common Knowledge, 3(2), 29–64. [PDF]