2. FEI-Based Arguments Against Capitalism

There is, at present, no overview of the arguments given for why FEI can condemn capitalism. This overview is organized historically, and focuses on four arguments from literature available in German and English that make the strongest philosophical claim on our attention.Footnote ⁸

FEI was first applied critically to capitalism by Hermann Cohen, founder of the Marburg school of neo-Kantianism, in a study of Kant’s aesthetics (Cohen Reference Cohen1889). For Cohen, FEI expresses the thought that ‘each individual is, as a moral person, the final end of all things’ (1889: 140). As such, FEI condemns the ‘statistical causality under which human beings, as values of labour, have languished’ (1889: 138). To treat producers with the attitude ‘If producers work for time t, then they will produce monetary value m’ ignores producers’ value as persons:

Cohen implies that, to treat producers as persons, the work they are offered must have a quality that serves a purpose beyond monetary value.

This claim rests on a premise made explicit in Ethics of the Pure Will (1904). In this work, Cohen asks what it would take to treat producers as persons, on the premise that ‘the vocation for morality cannot be fulfilled except on the basis of scientific insight’ (1904: 477–8). Scientific insight may alert agents to the occasions of a moral end, and so enable them to fulfil their moral vocation.Footnote ⁹ As persons ought to fulfil their moral vocation, then, by the principle of ought implies can, everyone must be able to partake in intellectual life. Our problem is that ‘man … has not come to partake in intellectual labour’ (Cohen Reference Cohen1904: 478). Producers are consigned to menial machine labour by the constraints of ‘market price’ (Cohen Reference Cohen1904: 305). If the nature of capitalism sustains general claims about the monetary value that the capitalist seeks by employing producers, and the menial type of labour and long hours this often requires, then, if this type and amount of labour hinders those who perform it from acquiring scientific insight (by exhaustion or dulling their minds), capitalism hinders (some) producers from fulfilling their moral vocation as persons. We can call this FEI-based argument against capitalism the division of labour argument.

Karl Vorländer first argues that Kant’s ethics provide a justification for socialism in Kant und der Sozialismus (1900). In this text, Vorländer claims that FEI formulates ‘the basic idea of socialism, the thought of community’ (1900: 7). FEI commands us to treat humanity in the person of even ‘the poorest day labourer’ as an end in itself (ibid.). Indeed, Vorländer holds against Kant:

This (as we might call it) community argument from FEI against capitalism has the structure: FEI expresses the thought of moral community; moral community presupposes universal political independence; and universal political independence presupposes universal economic independence, which capitalism denies to producers.

Consider the step from moral community to universal political independence. If we suppose that universal political independence, understood as universal rights of active citizenship, can guarantee for all the opportunity to assess and decide on principles that ought to be publicly declared and enforced, and that political life is thus, at least potentially (when universal political and economic independence are combined), a sphere in which everyone can develop the ability to judge and act on principles; then we may view universal political independence as a necessary preparationFootnote ¹¹ for all agents to cultivate the moral disposition that they must exercise in promoting one another’s permissible ends as members of a moral community.

Vorländer elaborates on the final part of the community argument in a slightly later text, Kant und Marx (1904):

This text permits two accounts of why universal active citizenship that is morality-preparing rests on universal economic independence. The first is that, if, in virtue of inequalities of private wealth, there is a lack of reciprocity between producers and possessors, then political life can be expected to cultivate acquiescence in an order by which possessors are enabled to exempt themselves from rules that producers must follow. Universal active citizenship will then not serve to prepare a moral disposition.

A second account is suggested by the thought that freedom requires each to serve ‘the whole’, and by Vorländer’s rejection of the ‘planless anarchy of value production’ (1904: 15). The idea would seem to be that, if separate units of production compete to maximize profits at each other’s expense, that will encourage a disposition to seek and bend public enforceable principles in one’s own favour. By implication, a moral disposition can only be enabled in a political community whose economic institutions allow citizens to pursue a common aim in their working lives. In the first version of the community argument, the problem with a capitalist economy is the lack of reciprocity, which might be possible to secure in an egalitarian market economy. In the second version, the problem is that any one person’s opportunity to work and acquire means of life remains under the control of an aggressively competitive sub-group, which a form of planning would have to replace.

Moving now to the Anglophone context, Harry van der Linden offers an institutional co-legislation argument. Van der Linden interprets FEI from within a broader account of Kant’s ethics as ‘a social ethics’, i.e. an ethics that includes ‘the demand that we actively seek the realization of ideal institutions’ (1988: 4). Ideal institutions are institutions that include ‘all’ participants as ‘legislative members’, and that aim ‘at the satisfaction of the ends’ of all their participants (van der Linden 1988: 31). A social ethics requires us to direct our efforts to realize a world in which all our institutions instantiate these features (van der Linden Reference Van der Linden1988: 163–4, Reference Van der Linden1994: 155). For if human beings can only do all they can to promote a goal if they ‘coordinate and optimize’ (van der Linden Reference Van der Linden1988: 163) their efforts by institutional means, the realization of a moral ideal among human beings must include the bringing into line of these institutions with this moral ideal.

FEI is a formula of a social ethics, for van der Linden, because humanity is a ‘task-setting idea’; ‘the idea of humanity does not refer to humanity as it now exists but to humanity as it ought to exist’ (1988: 29). Humanity only exists as it ought if institutions promote the highest good by making all participants legislative members and aiming as far as possible to satisfy their ends. For van der Linden, ‘to treat humanity in other persons as a means only is to act contrary to the demand that we should create a society of legislators who seek to promote one another’s personal ends’ (1988: 30). Van der Linden concludes: ‘[t]he cooperative fulfils this command, whereas the capitalist private enterprise violates it’ (1988: 225). The labour contract offered by a capitalist private enterprise ensures that the worker is not a co-legislator, but rather ‘his will is determined by another will – the will of the capital owner’, and thus ‘is treated as means only’ (van der Linden Reference Van der Linden1988: 225–6).

Finally, Allen Wood suggests an argument from FEI against capitalism that condemns playing on another person’s vulnerability for one’s own ends. For Wood, FEI requires us ‘to give the equal absolute worth of every human being its due’ (1995a: 316). Wood then adds that, ‘when a social order treats some people better and some worse in ways that they themselves regard as essential to their self-worth, there is a presumption, based on [FEI] itself, that this social order fails to respect the humanity of those who receive worse treatment’ (1995a: 316). A social order may fail in this regard if ‘untrammelled external freedoms in the economic sphere lead to large inequalities in wealth and status between human beings which are degrading to those placed in an inferior position’ (Wood Reference Wood1995a: 316; cf. 2014: 267, 296–8). In capitalism, workers are degraded by their bargaining interactions with capitalists in virtue of the fact that workers’ lack of means of production ensures that they ‘have no acceptable alternative’ (Wood Reference Wood1995b: 149) to working for a capitalist. The capitalist’s use of the worker’s vulnerability is degrading, because ‘[p]roper respect for others is violated when we treat their vulnerabilities as opportunities to advance our own interests or projects’ (Wood Reference Wood1995b: 150–1). ‘Kant’s principle that we must not treat others merely as means’ (1995b: 151) may account for this objection, Wood suggests. The presumption that FEI condemns capitalism can be rebutted only if such treatment is necessary to avoid conduct even more damaging to workers’ self-worth; and if capitalism avoids this conduct (cf. Wood Reference Wood1995a: 316).

3. Kant’s Never Merely as a Means Principle

An agent violates FEI by failing to treat humanity in a person not merely as a means, or by failing to treat humanity in a person as an end. One common feature of the above arguments against capitalism is that they all appeal to the Never Merely as a Means principle. This reflects a sound intuition. If one seeks to condemn capitalism by appeal to FEI, there is reason to focus on the Never Merely as a Means principle, at least as distinct from the Always as an End principle.

Kant’s statement of FEI reflects his distinction between strict (or narrow) duties, and wide duties (GMS, 4: 424; MS, 6: 390). A strict duty is one that an agent has no latitude in observing. A wide duty is one that an agent is afforded latitude in fulfilling: the agent may decide on the degree to which, and how, they fulfil a wide duty, for they may be obliged to limit its perfection by another duty. The Never Merely as a Means principle grounds strict duties: an agent is unconditionally necessitated to refrain from action that disrespects the status of humanity in a person as an end. The Always as an End principle grounds wide duties: an agent is not unconditionally necessitated to perform any specific action that respects the status of humanity in a person as an end.

Suppose that there is a type of conduct that falls under a wide duty that a type of agent essential to capitalism cannot perform, at least in a certain context. Its absence may not be sufficient to condemn capitalism. It may be that the agent cannot treat humanity in a person as an end in that respect while also fulfilling another duty, which capitalism enables them to fulfil. Its absence is then a permissible limitation of one duty by another. Capitalism is not thereby characterized as wrong.

It might be thought that the Always as an End principle can conclusively condemn capitalism by condemning conduct within it that consists in the adoption of a maxim never to fulfil a wide duty towards an agent or class of agent, e.g. a maxim never to come to the aid of workers in one’s employ. There are two problems here. First, this maxim is one that an agent has no latitude not to adopt. As such, it is condemned by the Never Merely as a Means principle. But in any case, such a maxim is not the sort of maxim that defines capitalism. A capitalist economy enables and constrains certain types of conduct. For instance, each unit of production is constrained by competition to make and reinvest as much money as it can, because competition attaches future disadvantages to non-accumulation. A capitalist economy cannot, however, constrain every type of conduct in which capitalists engage in relation to their workers. Expressed in terms of the characters of Dickens’s novel A Christmas Carol: even if competition constrains Scrooge to keep Cratchit’s wage down, Scrooge may decide, without compromising his firm’s profitability, to pay for Tiny Tim’s medical insurance with his own income.

It is also crucial to note that the Never Merely as a Means principle says: never treat humanity in a person merely as a means. For Kant, humanity is a set of rational capacities distinct both from animality (a living being’s capacities for self-preservation, propagation and community) and from personality (the capacity for moral action) (RGV, 6: 26–7). Most basically, humanity is the capacity for setting ends: ‘[t]he capacity to set oneself an end – any end whatsoever – is what characterizes humanity’ (MS, 6: 392). Humanity also includes any capacity whose exercise is a rational requirement of setting ends. It includes the capacity to order ends, and to select means to an end.

However, although this view of humanity has support in Kant’s writings, it is not the only one used to interpret FEI. Vorländer (Reference Vorländer1900: 7) and van der Linden (Reference Van der Linden1988: 29) speak in this connection of ‘the idea of humanity’. For them, the idea of humanity includes the institutional preconditions of what Kant refers to as the ‘perfection’ (MM, 6: 386) of one’s own humanity; that is, the cultivation of a virtuous disposition. Van der Linden, as we have seen, takes ‘the idea of humanity’ to include institutions that make all participants legislative members. Similarly, Vorländer says that ‘the idea of a moral kingdom or kingdom of ends’ that we are to realize, and that institutions are required to enable, is ‘constituted by the idea of humanity’, and that FEI rests on this idea (1900: 7). Yet if, for Kant, the ground of agents’ duties to perfect their humanity in their ‘deeds’ is to ‘be worthy of the humanity that dwells within’ them (MS, 6: 386–7), a social ethics based on FEI will emphasize that the humanity that ‘dwells within’ agents is the ground of their duty to perfect institutions. That differs from basing this duty in FEI interpreted as a command ‘to strive for humanity as it ought to be’ (van der Linden Reference Van der Linden1988: 31).

Further, the Never Merely as a Means principle commands us never to treat humanity in a person merely as a means. A person is a being with personality, i.e. the capacity to subject one’s conduct to the moral law, and act for duty’s sake (RGV, 6: 27). The Groundwork says: ‘morality is the condition under which alone a rational being can be an end in itself’ (GMS, 4: 435; cf. KpV, 5: 131–2; KU §83, 5: 431). If a capacity for moral action is the condition under which a rational being can be an end in itself, we can only be required to treat humanity in persons never merely as a means. FEI does not directly constrain action toward beings of a kind whose nature is such that their ends are ‘always wrong’ (Ameriks Reference Ameriks2003: 200), or whose nature is such that they can only comply with duty on heteronomous grounds. The relevance of this point for applying FEI beyond Kant will be made evident in section 5.

What remains clear in any case is that the Never Merely as a Means principle requires us never to disrespect the worth of humanity in a person that gives it the moral status of always being an end. What gives humanity in a person this moral status, Kant claims, in arguing for FEI, is that humanity in a person is ‘something the existence of which in itself has an absolute worth’ (GMS, 4: 428). We therefore require an analysis of absolute worth.

The Groundwork leads us to believe that absolute worth contrasts with ‘relative’ (GMS, 4: 428) worth. One way to begin to think about absolute worth, therefore, is suggested by Scanlon: ‘seeing something as an end in itself is seeing it as having value that is not derived from any other source’ (2008: 92). By contrast, ‘to regard something merely as a means is to see it as having no value except for what may be conferred on it’ by something else (Scanlon Reference Scanlon2008: 92). This suggests the following analysis: something exists with absolute worth if it exists with non-derivative worth.

This suggestion is not yet adequate, however. If X has some non-derivative worth, but X’s non-derivative worth cannot equal Y’s non-derivative worth, then, as a thing of lesser value than Y, X may be treated as a mere means with respect to Y’s worth. So, to say that X has non-derivative worth does not imply that X ought always to be treated as an end. What needs to be added is the notion of transcendent worth. Y has transcendent worth if no amount of anything that is not Y can have a value equal to Y (cf. Kerstein Reference Kerstein2013: 129). If Y’s non-derivative worth is a transcendent worth, Y ought always to be treated as an end.

By itself, transcendent worth does not suffice for absolute worth, for transcendent worth may still be derivative. Y has transcendent but derivative worth if Y owes its transcendent worth to an external source, Z. If this external source is cut off, Y may continue to exist, just not with transcendent worth. Something of transcendent worth only always exists with absolute worth (and so must always be treated as an end) if it is also its own source of worth, i.e. if it has non-derivative worth. On this analysis, the Never Merely as a Means principle requires us never to disrespect the non-derivative and transcendent worth of humanity in a person. This finally puts us in a position to gain more clarity regarding the problem of applying FEI to the case of capitalism.

4. Applying FEI: Some General Considerations

First, it is important to consider the general form of argument by which a duty may be derived from FEI. A good place to start is Allen Wood’s claim that every derivation of a duty from FEI ‘depends on an intermediate premise, logically independent’ (1995a: 313) of FEI itself. In Wood’s view, if FEI requires us to treat humanity in a person always as an end and never merely as a means, then, to conclude that FEI imposes a duty on us (say, not to y), it is necessary to defend a premise that says: to y is to treat humanity in a person merely as a means, or to fail to treat it as an end. An ‘intermediate premise’ of this form is said to be logically independent of FEI because it is ‘possible without contradiction’ (Wood Reference Wood1995a: 314) to deny it while affirming FEI.

One drawback of Wood’s view, however, is that, if rational nature is something about which it is possible to have knowledge (Wood Reference Wood1995a: 315), then it ought not to be possible, without contradiction, to deny that FEI imposes any given duty on us, while affirming FEI. Some duties on us derived from FEI ought to be incontrovertible. Our affirmation of FEI ought to commit us to at least some duties, even if other duties remain controversial.

Fortunately, it is not necessary, while affirming FEI, to deny that FEI imposes any given duty on us, in order to hold that every derivation of a duty from FEI rests on a premise logically independent of FEI. For the form of this premise may be altered to: ‘in the context of x, to y is to treat humanity in a person merely as a means, or to fail to treat it as an end’. Any premise of this form can be used, with FEI, to derive a duty (not to y). A premise of this form is also logically independent of FEI, for one may conceive of a hypothetical world in which x is not a possible context, and so deny that the premise is relevant to moral deliberation in that world, while holding that FEI remains valid for it. One may, for example, conceive of a one agent world in which no context in which an agent invites another person to act is possible, and so deny that the avoidance of false promising is a duty in it (there being no duty not to do something that is causally impossible), while holding that FEI remains valid in that world. As a premise of the above form can be used, together with FEI, to derive a duty, it is not necessary, while affirming FEI, to deny that FEI imposes any given duty on us, in order to hold that every derivation of a duty from FEI rests on an intermediate premise logically independent of FEI.

The form of argument for deriving a duty from the Never Merely as a Means principle can be illustrated with the Groundwork’s example of a lying promise to obtain money. Kant says that someone who makes a false repayment promise uses another merely as a means, for ‘the one I want to use for my purposes by such a promise cannot possibly agree to my way of proceeding with him and thus [also] himself contain the end of this action’ (GMS, 4: 429–30). The ‘thus’ in this statement is crucial. If A makes a false repayment promise to B, B cannot will to contribute to A’s end by virtue of being given a chance to do so by A. A fails to give B this chance, by inviting B to act on reasoning about A’s end in respect of B that A must believe is false (namely: ‘if I give A this money, A will repay me’). As it is part of A’s end to get B to act as B would not otherwise act, A is obliged to invite B to act on a sound piece of reasoning about A’s end in respect of B. To invite B to act on reasoning about A’s end in respect of B that A must believe is false (that is, to make a false promise) is to do the contrary of this.

The argument can be represented as follows:

1. (1) Never treat humanity in a person merely as a means.

2. (2) In the context of inviting another person to act as they otherwise would not, it treats their humanity merely as a means not to give them the chance to act on reasoning about one’s end in respect of them that one believes is true.

   Therefore:

3. (3) It is a strict duty not to make a false promise to another person.

(1) is Kant’s Never Merely as a Means principle. (2) is a premise that concretizes (1) in a certain context. (3) states a specific duty. If (1) and (2) are true, (3) is true. On the analysis defended in section 3, the argument for (2) is to take the form of arguing that the conduct in (2) disrespects the non-derivative, transcendent worth of humanity in a person.

One general point to note is that the contextualizing premise (2) must relate FEI to the act and end of a piece of conduct. FEI begins ‘so act …’, and commands us always to treat humanity in a person ‘as an end’. It is impossible to judge whether a piece of conduct treats humanity in a person as an end, and whether it treats it merely as a means, without subsuming it under a contextualizing premise that relates FEI to an end to which an act is oriented.

The idea that every derivation of a duty from FEI rests on a contextualizing premise of this kind raises a number of issues. Two may be mentioned here.

First, the description of conduct to choose when subsuming it under a contextualizing premise must state the agent’s end, or reason for action. It does not suffice for it to report an action’s consequences. Suppose A defends herself against an unprovoked violent attack by B. In applying FEI to A’s conduct, it would be mistaken to look for a contextualizing premise under which to subsume a description such as: A injures B, and thereby damages B’s humanity. This description does not state A’s end, to defend herself from B.

Second, if the contextualizing premise relates FEI to an act and end, a charge of treating humanity in a person merely as a means must rest entirely on the conduct of the agent(s) against whom the charge is levelled. There must be an agent (or associating agents) against whom the charge can be levelled, because only agents can act on an end; and third-party influences that impact on the addressee’s overall situation are not relevant just by virtue of that, for they are caused by parties unrelated to the agent(s) whose conduct is under evaluation.

Take A’s false repayment promise to B. A cannot be saved from a charge of treating B merely as a means by the fortuitous fact that B may know what is going on from C, who informs B of A’s plan to deceive B, and so puts B in a position to choose whether or not to contribute to A’s end, for example, after B considers whether they want to ‘revel in the demise’ (Kerstein Reference Kerstein2013: 61) of A’s reputation as a result of A achieving A’s end of obtaining money from B by a false promise.

Contrariwise, if B is falsely convinced by a third party, C, that what A proposed to B differs from what A said to B, for example, C convinces B that A spoke to B in code, that does not mean that A treats B’s humanity merely as a means, even if the confusion sown by C leaves B unable to will to contribute to A’s end. Similarly, if B is powerless to decline A’s offer by virtue of conduct by a third party, C, that A is unaware of (C has, say, hypnotized B to say ‘yes’ whenever B wants to say ‘no’), that does not mean that A treats B’s humanity merely as a means by making B an offer.

In short, the Never Merely as a Means principle only condemns conduct, rather than its addressee’s overall situation; and it need not condemn conduct whose efficacy depends on its addressee’s ignorance or powerlessness. For in neither of the latter cases does A treat B’s humanity merely as a means if it is also assumed that B would certainly decline A’s offer if B knew that A was not speaking in code, or if C had not hypnotized B. If the ‘nature and efficacy’ (Korsgaard Reference Korsgaard1996: 139; emphasis added) of conduct depends on its addressee’s ignorance or powerlessness, then it will count as treating her merely as a means.

Two related points pertain to applying FEI to condemn capitalism. First, when applying FEI to condemn capitalism, the description of the type of conduct essential to capitalism that is to be subsumed under a contextualizing premise must state its act and end (not a mere consequence). Second, any charge that an agent (or associating agents) treats the humanity in a worker’s person merely as a means must rest entirely on the conduct of the agent(s) against whom the charge is levelled. This agent may in principle be a capitalist, an association of capitalists or a capitalism-guaranteeing government. (Depending on the conception of capitalism, a capitalist is a member of a class of private owners of means of production or a representative of a profit-maximizing unit of production.) A capitalist’s end in hiring a worker is to obtain the power to put the worker to work on the agreed terms. A capitalist’s end in putting a hired worker to work on the agreed terms is to appropriate the product. If FEI is to condemn capitalism, a duty must be derived from FEI’s Never Merely as a Means principle, by the form of argument outlined above, that either prohibits one of these two types of conduct, or prohibits a type of conduct (including by an association of capitalists or a capitalism-guaranteeing government) that is a necessary condition of either of them.

5. Applying FEI beyond Kant

I begin by noting that a Kantian might use FEI to condemn conduct that Kant did not use FEI to condemn by adopting one or more of the following strategies:Footnote ¹²

1. (1) Revise a value criterion in Kant’s a priori argument for FEI as a categorical imperative, to strengthen the justification of FEI as a formula of a moral law whose principle is autonomy (i.e. argue that a value criterion other than ‘existing with absolute worth’ qualifies something as an obligatory end, or reason for acting).Footnote ¹³

2. (2) Revise Kant’s statement of FEI, to better capture the value of rational nature to which FEI is to give expression (e.g. to treat humanity in a being always as an end, never merely as a means).

3. (3) Affirm a contextualizing premise that Kant did not affirm.

4. (4) Subsume under a contextualizing premise a description of conduct that Kant did not consider.

As we shall see, the arguments from FEI against capitalism presented in section 2 can all be understood as attempts to pursue one or more of these strategies. (They all pursue strategies (3) and (4).) The aim of the following evaluation is to show that, even if plausible versions of these strategies are permitted, none of the arguments presented above allows FEI to condemn capitalism. To anticipate: the first three arguments rewrite FEI so as to be unrecognizable as arguments from an obligatory-ends-based formula of a moral law whose principle is autonomy; while the vulnerability argument relies on a description of conduct in capitalism to which FEI cannot be applied.

6. Evaluation of the Arguments