A Neo-Kantian foundation of Corporate Social Responsibility

Abstract

‘Corporate Social Responsibility (CSR) is conceptualized in many ways. We argue that one cannot be indifferent about the issue of its conceptualization. In terms of methodology, our position is that any conceptual discussion must embed CSR in political theory. With regard to substance, we link up with the discussion on whether CSR must be defined on the basis of a tripartite or a quadripartite division of business responsibilities. We share A.B. Carroll’s intuition that a quadripartite division is called for as a basis to define CSR. However, defending the quadripartite division of business responsibilities requires that the distinctions between economic, legal, ethical and discretionary business responsibilities be made intelligible. Carroll’s account is defective in this respect. We argue that contemporary Neo-Kantian political ethics is able to make sense of these distinctions, because of its specific interpretation of liberalism. Interestingly, from a conventional liberal perspective this interpretation of liberalism is atypical, as it extends public morality beyond the domain of the law.

1 Introduction

(P)erhaps the most difficult issue concerning ‘corporate social responsibility’ is that…(o)ne searches in vain for any clear and generally accepted moral principle that would impose on business an obligation to work for social betterment…the intractability of (this) central issue has…posed the dreadful possibilities that the debate…would either continue indefinitely with little prospect of final resolution or that it would simply exhaust itself and collapse as a viable legitimate question (Frederick 2006: 39)).

There is something peculiar about the conceptualization of CSR as dealt with in business ethics and business and society theory.¹ While there is a great proliferation of definitions (MacWilliams et al. 2006), there is not as much discussion on definitional issues as one would expect—at least from a philosophical point of view. Many authors endorse one definition and seem to be indifferent about other authors using other definitions. Often they simply state that this is inevitable when concepts are complex or dynamic (see Van Marrewijk 2003).

Still, even if CSR is an “essentially contested concept” (Gallie 1956), indifference is not a proper response to a proliferation of conceptualizations. Definitional issues matter, if only because of their practical impact. This can be illustrated by explicating the practical consequences of making a choice in relation to the crucial issue of whether CSR must be conceived of as a moral duty and if so, what such a conceptualization entails. Experts in business ethics are deeply divided over this issue. Some, such as Carroll (1979; Carroll and Buchholtz 2006), hold that CSR is indeed a moral duty. Others, such as Manne and Wallich (1972) and the Dutch Social Economic Council (2000), stress that CSR must be conceived of as a completely voluntary endeavour. A market actor can decide upon it at will, just as he can decide whether or not to sponsor a sports club. Consequently, it cannot be conceived of as a moral duty, as these are not voluntary in this sense. Choosing among these alternatives has practical consequences, for example, in relation to the position of NGOs in the political and economic arenas and in terms of how the general public view CSR. If CSR is a practice that corporations can decide upon at will, then any steps taken against corporations that refuse to act in a socially responsible manner must be different in nature from a situation in which CSR is considered a moral duty. For example, it does not seem proper to morally frown upon or even morally punish actors for performing (or not performing) certain actions that are completely voluntary, let alone to call upon them to perform such actions collectively. Furthermore, if CSR constitutes a practice that an actor may decide upon at will, there is no contradiction whatsoever between CSR and the prospect of gain. Consequently, other parties—such as the government—may try to entice that actor by capitalizing on its self-interest, as is often the case in latter-day public policy aiming at strengthening CSR. If CSR is conceived of as a moral duty, this matter will not be as clear cut, to say the least. Moreover, if CSR is a moral duty, this means that the businessperson is bound to perceive it differently than if it were an act which can be decided upon at will. He cannot dismiss it because it is unprofitable. After all, even if the businessperson’s² attitude towards his moral duty is one of reluctance, a duty cannot be rejected for such reasons.

It is our aim in this paper to contribute to the clarification of the concept of CSR, in particular its moral nature. We organize our argument by connecting it to an established opposition within business ethics: the discussion as to whether a definition of CSR should be based upon a tripartite or a quadripartite division of business responsibilities. In a tripartite division of business responsibilities, a distinction is commonly made between economic, legal and social responsibilities. CSR is thereby identified with the social responsibilities. We can also base our conceptualization of CSR on a quadripartite division of business responsibilities. Accordingly, a distinction can be made between economic, legal, ethical and discretionary responsibilities. In that case, CSR can be identified with the ethical responsibilities of the businessperson.³

The shift from a tripartite to a quadripartite categorization of business responsibilities is vital for our understanding of CSR, in particular in relation to its moral nature (i.e. the issue of whether it can be conceived of as a moral duty). In a tripartite division of business responsibilities, all acts of CSR are lumped together in a category that includes acts of corporate voluntarism and philanthropy. Tripartite CSR is therefore conceived of as a practice which corporations are free to decide upon at will (even if deciding against it may involve costs due to the reputation mechanism). If we conceptualize business responsibilities in terms of a quadripartite model, we are able to detach at least part of the extra-legal responsibilities from subjective and free choice and conceive of them in some other way. Accordingly, in a quadripartite categorization of business responsibilities, we need no longer conceive of CSR in its entirety as a purely voluntary practice. We have carved out conceptual space for another kind of requirement. Its acknowledgment no longer has to be a matter of subjective and free choice on the part of the actor involved. The requirement can have a non-voluntary status of some sort.

A.B. Carroll is the leading latter-day protagonist of the quadripartite division of business responsibilities. The conceptual moves described above occur in his thinking. Voluntarism and philanthropy are classed under the discretionary responsibilities; a category that is explicitly distinguished from the ethical responsibilities. What is more, Carroll assumes the binding force of the ethical duties to be of a different kind. “(I)t deals…with ‘what ought to be’ or ‘what ought not to be’ in terms of business practices” (Carroll and Buchholtz 2006: 175). We concur with Carroll’s insistence that a definition of CSR must be based upon a quadripartite division of business responsibilities. However, we believe that there are serious philosophical problems with the way in which he conceives of them. Carroll’s formulation of these distinctions fails to make them intelligible from a systematic point of view or justifiable. Needless to say, this seriously undermines his account. The main contribution of this paper is to show that Neo-Kantian political theory can account for a quadripartite division of business responsibilities. It can justify and make intelligible the distinctions between law, morality and discretionary actions. In doing so, it can also conceive of CSR as a moral duty. We thereby classify Neo-Kantian political ethics as a school within the broad current of liberalism. Surprisingly, Neo-Kantian thinking is only able to serve as shell for CSR as a result of some aspects that may well be regarded as almost atypical, when seen from the perspective of liberalism. We will also show that the Neo-Kantian account of these distinctions differs from Carroll’s division in a number of important respects.

Our second contribution is methodological. As we see it, the problem with current discussions on CSR within business ethics is not just that there is no “generally accepted moral principle”, as Frederick states in the introductory quotation. The problem is more extensive. It also encompasses the issue of what sort of argument should be used to settle the issue as to the tripartite or quadripartite division of business responsibilities. That is to say, there also is hardly any reflection—and thus conscious consensus—as to the method by which this generally accepted principle is to be decided upon. How can we set about defending the quadripartite division as the proper one? In this regard, we will argue that discussions on the conceptualization of CSR ought to be conducted and decided upon in the context of political ethics or political theory.

We are well aware that our contributions are merely a stepping stone in the stream of debate on the proper conceptualization of CSR. We do not even set out to settle the specific discussion between tripartite and quadripartite conceptualization. In order to settle that discussion, it would be necessary to show that the tripartite division of business responsibilities cannot be made intelligible and justified within liberalism. Alternatively we would have to show that the Neo-Kantian school is superior to all alternatives (within liberalism). Both tasks fall outside the scope of this paper. Still, we think our contribution is relevant in that it supplies the necessary building blocks to launch any one of these further investigations.

We will start with a critical examination of Carroll’s quadripartite categorization, leaving aside the category of economic responsibilities, as these would confront us with specific problems not related to our subject (Section 2). Next, we will argue that any justification of CSR must contextualize the phenomenon in the context of a political theory (Section 3). Accordingly, we explicate Neo-Kantian political ethics and show that it has at its disposal exactly the correct distinctions and other characteristics to make the quadripartite division of business responsibilities intelligible. We will develop the Neo-Kantian perspective in relation to other strands of liberalism, thereby substantiating the claim that it is a number of rather atypical aspects of Neo-Kantian political ethics upon which we need to rely (Section 4). We will subsequently return to Carroll’s thinking and briefly indicate the important distinctions between his conceptualization and the Neo-Kantian perspective (Section 5). The last section not only contains a conclusion, but also features a tentative basis from which to argue that a quadripartite division of business responsibilities is more sensible than a tripartite division.

2 Carroll’s Pyramid and its Conceptual Difficulties

Many conceptualizations of CSR are based on a framework that distinguishes between three types of business responsibilities. One example is the approach by Post, Lawrence and Weber. These authors (Post et al. 1999: 70) categorize the responsibilities of business as economic, legal or social. On the basis of this categorization they identify CSR with the social responsibilities.

One of the main challengers of this tripartite division is A.B. Carroll, who in 1979 launched his quadripartite categorization of business responsibilities. According to Carroll:

…previous definitions had alluded to businesses’ responsibility to make a profit, obey the law, and ‘go beyond’ these activities….I observed that to be complete, the definition had to embrace a full range of responsibilities of business to society….Therefore, I offered the following definition: the…responsibility of business encompasses the economic, legal, ethical, and discretionary expectations that society has at a given point in time (Carroll and Buchholtz 2006: 500).⁴

The critical step that Carroll takes, and that results in his famous CSR pyramid (see Fig. 1), is to split the category of the social responsibilities in two. Thus he distinguishes between economic responsibilities, legal responsibilities, ethical (or moral) responsibilities and discretionary responsibilities.⁵ In this way, Carroll creates the possibility of a more qualified approach to CSR, as it can be related to the ethical responsibilities. He thereby explicitly states that these responsibilities do not represent any hierarchical order, even if the shape of the pyramid suggests otherwise.

Fig. 1

Carroll’s pyramid of corporate social responsibility. See Schwartz and Carroll 2003: 504

In the latest version of his handbook Business and Society, Carroll identifies the four categories of responsibility as follows. The economic responsibility of a market actor is to “produce goods and services that society wants and sell them at fair prices” (Carroll and Buchholtz 2006: 35). The legal responsibilities are implicated by the fact that society has to formulate “ground rules—the laws—under which business is expected to operate….It is business responsibility to society to comply with these laws” (idem). According to Carroll, ethical responsibilities come in because laws are “not adequate (.E)thical responsibilities embrace those activities and practices that are expected or prohibited by societal members even though they are not codified in law” (idem: 37). They embody the full scope of “norms, standards and expectations that…(…people…)…regard as just, fair and in keeping with the respect for or protection of stakeholders’ moral rights” (idem). He even goes so far as to speak of ethical responsibilities as duties on occasion (idem: 173). Carroll strongly associates this category with “avoiding harm” and “doing what is just and fair”. The last category, labelled discretionary or philanthropic responsibilities “are viewed (as responsibilities) because they reflect current expectations of business to the public” (idem: 37). According to Carroll, these responsibilities are voluntary, “guided only by business’s desire to engage in social activities that are not mandated, not required by law, and not generally expected of business in the ethical sense” (idem). Examples of philanthropic responsibilities are occasional donations to a local sports club or a management initiative to take a day off to clean out or renovate a children’s farm.

Although we share the intuition that CSR presupposes a quadripartite division of responsibilities, we also maintain that Carroll’s conceptualization poses significant problems. In Carroll’s account of it, the fundamental distinctions between legal, moral and philanthropic responsibilities do not stand up to scrutiny. First, there are some philosophical problems with the way Carroll clarifies the distinction between legal and moral duties. Carroll (Carroll and Buchholtz 2006: 37 and 177) makes sense of this distinction by claiming that often legal laws “are not adequate”. Therefore a supplement of extra-legal rules is needed. He also accounts for the distinction by claiming that ethical rules are “ill defined and continually under public scrutiny”. Both these claims are mistaken. Many moral rules are not ill defined, certainly not in comparison to legal rules. In many cases, these also involve concepts like “reasonable doubt”, “reasonable risk” or “reasonable response”, which need to be clarified by jurisprudence (Buchanan 1987: 560–561). Examples of moral rules that are at least as clear are the rescue principle or the principle that one ought not to steal. Furthermore, it is questionable whether we should conceptualize morality as an addition to the law. There are many circumstances in which we seem to conceptualize their relation differently, such as a case in which a person who is known to have committed a grievous assault is let off because his lawyer finds a legal loophole that forces the court to acquit him on a technicality. Without falling into a contradiction, we can say in such instances that this person acted in a way that was morally wrong, even if he is legally acquitted. Should we therefore not conceive of law and morality as independent systems with a fundamentally different ground, even if some legal and moral laws are identical? In other words, it is often said that moral responsibilities are “above the law”. Carroll (1999: 177) acknowledges this. As we see it, this means that morality is fundamentally beyond the grasp of the law. Legally speaking, it is something that is fundamentally voluntary. Now, if morality is merely additional to the law, this puts it “above the law” in a very different respect. In that case, morality and the law are not fundamentally distinct. It seems that we must therefore conceive of morality as “proto-law”: law that is not codified due only to a host of contingent pragmatic reasons.⁶

Interestingly, in his later work Carroll seems to revise his thinking on the relation between morality and the law:

(W)e said that ethical behaviour is typically thought to reside above behaviour required by law. This is the generally accepted view of ethics. We should make it clear, however, that in many respects the law and ethics overlap…(…)…Both law and ethics have to do with what is deemed appropriate and acceptable, but the law reflects society’s codified ethics (Carroll and Buchholtz 2006: 177).

There are two problems with this alternative view. First, it still cannot explain the fundamentally voluntary nature of morality vis-à-vis the law. If law is “codified ethics” and law is enforceable, why then is it principally impossible to enforce morality if it is in fact “uncodified law”? Second, the alternative view is mistaken, as examples such as abortion or the death penalty show. Both these practices are legally permissible, yet many people have their doubts as to the moral rectitude of either or both. The point here is not so much that the law and ethics only partly overlap. The point is that the overlap between law and ethics only concerns concrete material principles, such as “do not steal” or “do not kill”. Both fundamentally differ as to their meaning. Examples such as abortion and the death penalty demonstrate that the law is not an authority on moral issues (even if we—from a moral point of view—are required to respect the law at all times). The reverse also holds. A person’s moral rectitude is irrelevant in the context of the law. Hence there is a fundamental problem in conceiving of law as codified ethics (i.e. morality as uncodified law).

Let us now turn to the distinction between moral responsibilities and philanthropic responsibilities. A first problem with Carroll’s account is that he describes both in terms of society’s expectations (Carroll and Buchholtz 2006: 37; Carroll 1979: 500). As we see it, it is somewhat strange to speak of ethical responsibilities as ’expectations’. An expectation is something that we hope market actors will do; that we have reason to believe that they will do. We may even be disappointed or surprised if they fail to act in this way. But that’s the end of it. A moral duty is something different. We hold that a person must obey it (even if we sometimes do not expect her to).

Carroll tries to make sense of the distinction between moral responsibilities and philanthropy in two ways. The first way is by saying that philanthropic expectations are “voluntary” (Carroll and Buchholtz 2006: 37; Carroll 1979: 500), whereas ethical responsibilities must be compared to legal responsibilities, in that both “have to do with what is deemed appropriate and acceptable” (Carroll and Buchholtz 2006: 177). In our view, this again turns morality into something akin to proto-law and we have already established that this is mistaken. The second way in which he tries to clarify the distinction is by saying that philanthropic expectations are “not generally expected in an ethical sense” (idem: 177). As we see it, this only begs the question as to why such expectations are specific to morality.

In summary, although it is crucial for Carroll’s project to distinguish law, morality and discretionary responsibilities, he cannot fully account for these distinctions. Consequently, the crucial category of “ethical responsibilities” remains somewhat vague. In what sense are these obligatory? What claim does one person have over another person with their own ethical responsibility?⁷

3 How to Justify any Conceptualization of CSR

It is not our objective to end all discussions on the definition of CSR. In a field as contested as the conceptualization of CSR, disagreement is inherent. However, our ambitions extend beyond simply providing yet another justification of the quadripartite categorization of business responsibilities. We want to resolve one aspect definitively. Our aim is to justify the quadripartite division of business responsibilities in the way that it ought to be justified. This raises a key methodological question: what sort of demonstration or argument can be used to sustain any particular conceptualization of CSR? This question is not often asked in discussions on CSR. But, if there is any hope of making progress in the discussion on the conceptualization of CSR, it is imperative to reflect on it.

Once the question has been posed, the answer seems obvious. CSR is in fact a normative theory relating to the role that market actors have to play within the context of a free market economy, an economy that is itself embedded in a differentiated liberal society. It describes the moral duties and discretionary powers of a market actor, given a larger theoretical context of the functioning and the value of the free market, and given an even broader context in which this free market itself makes sense. This implies that a proper conceptualization of CSR should be offered within the context of an overarching political theory. After all, justifying the organization of the free market as an institution is a task for political philosophy, broadly conceived (see also: Peter 2004; Scherer and Palazzo 2007). This position is consistent with the fact that the fiercest and most troubling attacks upon CSR are indeed launched from the perspective of political theory (Baumol 1975; Baumol and Batey Blackman 1991; Homann 1994). These criticisms argue that CSR does not make any sense in a liberal free-market system.

Of course, the line between “moral theory” and “political theory” is not clear cut. Moral theory is sometimes defined so broadly that it completely encompasses political theory. Our intention in stating that CSR has to be justified within the context of political theory is that—even if all market actors were humans—CSR is not about actors as natural persons. Nor is it about natural persons having to determine their mutual relationships. In that sense it falls outside of moral theory proper (i.e. in the narrow definition of the term). CSR is about the duties of market actors in a particular role. The particularities of this role (such as the notion that one is allowed to concentrate on increasing personal gain) acquire meaning in the context of the free market in its relation to society, the state, democracy and the values of modern Western society. It is this embeddedness that makes it necessary to justify CSR in the context of political theory.

4 Neo-Kantian Thinking as a Foundation for CSR

As it is our goal to see whether the quadripartite conceptualization of business responsibilities can be made intelligible, we need to investigate whether the distinction between legal responsibilities, ethical responsibilities and discretionary responsibilities as fundamentally distinct categories can be made intelligible and be justified within a political theory. In order to find at least one feasible candidate to fulfil this task, it is relevant to observe that discussions on CSR take place within the context of a free market society. At the very least, this type of society values freedom and autonomy as basic values and advocates economic reproduction by means of the free market. The free market is also valued as a basic social institution because of its support in realizing freedom. These facts focus the search for a suitable candidate. It strikes us as a reasonable endeavour to embed CSR in the context of liberalism, conceived of as a broad school of thought that includes the likes of Mill, Kant, Rawls, Friedman, Hayek and Nozick. After all, liberalism generally embraces the (concept of a) free market.

In today’s political ethical landscape, two variants of liberalism are customarily distinguished: a libertarian school and a liberal democratic school.⁸ As our object is only to determine whether there is at least one political theory that can make sense of the distinctions between legal, moral and discretionary responsibilities, we need not make a fundamental choice between these strands. We simply have to determine which theory is the most promising, without having to claim at the same time that other theories are fundamentally ill-suited to accommodating the quadripartite division of business responsibilities. Given this condition, we opt for the liberal strand. Libertarian thinking is less promising, since it has a tendency to distinguish only two types of requirements: those that are dutiful and enforceable by law at the same time and those that are beyond duty (supererogatory) and left to any person’s subjective and free choice (Heyd 1982). Within this context, justifying the distinctions between law, moral duty and discretionary duties seems to hold little promise.

Within the liberal democratic school, we make a more specific choice for contemporary Neo-Kantian thinking as inspired by thinkers such as T. Scanlon 1998), Paul Guyer (2000a,b), Barbara Herman (1993a, 1997) and Onora O’ Neill (1986, 1989, 1993), Marcia Baron (1995) and Andrews Reath (1997a,b). Again, the reasoning behind this more specific choice is not to exclude any particular liberal democratic strand. Our aim is to find at least one political theory to make sense of the quadripartite division of business responsibilities. Still, two reservations are in order. Neo-Kantian political ethics can be divided into one variant that follows Kant in conceiving of liberalism as a comprehensive theory of the good and another—specifically Rawlsian account (1993)—that sets out to conceptualize liberalism as a ‘free standing’ political conception, one that is not dependent on a specific comprehensive theory for its justification. We do not want to get into this discussion and are content to simply acknowledge that we conceive of our account as a comprehensive one. Second, there is the issue of the distinction between Kantian and Neo-Kantian thinking. Some Neo-Kantian liberal democrats, such as Rawls, Scanlon and Reath, explicitly distance their own political ethical thinking from the Kantian position. Others are often silent when it comes to the possible distance between their views and those of Kant as a historical figure. This raises the issue what Kant “really meant”. Again, we do not want to become enmeshed in this discussion, whose relevance to our argument is only indirect. We will therefore provide a fairly standard—if not traditional—interpretation of Kant and attribute anything special to the Neo-Kantians involved. Since this paper’s main interest lies in the relation between CSR and contemporary political ethics, this strikes us as the best way to advance (even if some commentators might claim that we have credited Neo-Kantians with insights already foreseen by Kant himself).

“Human freedom” is the core value of liberal-democratic thinking. Naturally, it is not the only value. Equality is also considered to be important (Kymlicka 1990), as are justice, (mutual) respect and (minimum) solidarity. Within the Neo-Kantian account, freedom must be understood as autonomy. Freedom is not psychological freedom and it is not—or not primarily—the right and ability to make subjective and free choices. A free person is a self-governing person, capable of acknowledging the imperative nature of moral principles in the face of opposing inclinations (driven for example by self-interest or a desire for honour). Furthermore, an autonomous person has succeeded in cultivating her own character in such a way that the acknowledgment of morality is not—always or only—experienced as a sacrifice. A free person interprets autonomy as an end in itself.

The centrality of autonomy as an end indicates that the Neo-Kantian perspective is a moral perspective through and through. This is also reflected by the specific way in which it handles a fundamental problem of political theory: by what rules or principles should people live together? In thinking on the type of reasons that can be of importance in answering this question, some theorists only allow prudent reasons (Gauthier 1986; see also Schneewind 1998 and O’Neill 1996). Others maintain that any argument that does not anchor at least some rules or principles in man’s status as a moral being, will result in a seriously distorted political theory. Neo-Kantians firmly agree with this latter position. In this respect, they follow in Kant’s footsteps. Kant also radically conceived of man as both a moral and a reasonable (Vernunftig) being. He maintained that reasonable man has a moral duty to leave the state of nature, because freedom will only be possible in the "civil state" i.e. in the circumstance in which a people have constituted a state that can establish and enforce the fundamental rules guiding human interaction. “Anyone willing to remain in a pre-civil state is not just an irrational noncooperator, but thereby does wrong in the highest degree” (Pippin 417).

We will further explain Neo-Kantian thinking by investigating the relationship between three important issues of any political theory. These are:

• (a) Right, the sum of laws minimally needed to constitute a society of equally free citizens (Kant 1797: 23–24). These laws must be enforceable by the state in order to guarantee everyone’s freedom.

• (b) The full set of socially acknowledged general moral rules and principles requisite for the regulation of behaviour (to be abbreviated as “the full set of general principles”).

• (c) The pursuit of moral virtuousness by the individual person. (to be abbreviated as “the individual pursuit of virtue”).

Some within the broad school of liberalism are perhaps prone to think that Right and “the full set of general principles” are identical in a liberal society. It is therefore important to stress that this need not necessarily be the case. For example, Kant argues that any possible society of “rational beings” (devils, angels and humans) stands in need of Right. Freedom is not possible without a minimum set of enforceable rules in place (Höffe 1992). However, at least angels and humans would not find it rational to live in a society in which only Right reigned. The “full set of general principles” would be broader. Humans would, for example, include a duty of mutual aid. Angels would at least add a duty of being kind to each other and—possibly—to humans (Herman 1993a,b: 122). Neither of these duties are accounted for in terms of Right. That is to say, the fact that Neo-Kantian thinking is moral through and through is not merely an interesting contingent fact. It is crucial in this context, since it urges the Neo-Kantians to make distinctions that other strands within the school of liberalism do not necessarily have to make.

Of course, our reason for focusing our discussion of Neo-Kantian political theory on the relationships between (a) Right, (b) the full set general principles, and (c) individual pursuit of virtue is the strong resemblance between these categories and the distinctions needed for a quadripartite division of business responsibilities. For the purposes of clarification, we will present our Neo-Kantian account of these relationships in the form of an evolving perspective from (1) a common liberal perspective (2) a standard Kantian view to (3) the Neo-Kantian view. Our argument is that only Neo-Kantian theory is able to fundamentally distinguish Right, the full set of general principles and the individual pursuit of virtue.

• 1. The standard liberal interpretation of these relations is in fact a response to conservatism, i.e. a liberal perception of conservatism. According to this perception, conservatism maintains that none of the distinctions are fundamental in nature. The pursuit of virtue by the individual is believed to be a collective concern. Accordingly, we cannot fundamentally distinguish between the individual pursuit of virtue and the full set of general principles. Again, as the state is the representative of the people, it has a right and a duty to uphold the full set of general principles. Thus there is only a thin line between right and morality. Right (legality) is “embodied morality” (see Postema 1994).⁹

Liberalism passionately rejects the conservative view of the relation between Right and the individual pursuit of virtue. All liberal theories remove the individual pursuit of virtue from the legitimate exercise of power by the state. However, liberalism commonly concurs with conservative thinking in at least one respect: it does not make a fundamental distinction between Right and the full set of general principles. These two aspects are frequently assimilated (Postema 1994). This is often but not always done implicitly. It can be argued that liberalism, as it developed in the twentieth century, has struggled to free the individual from a smothering social morality that did not allow for the celebration of all kinds of different lifestyles, by stressing that there are no relevant general rules or principles outside Right. Consequently, liberalism commonly is not particularly well suited as a shell for a quadripartite division of business responsibilities.

The proposition that liberalism does not commonly—or necessarily—distinguish between the full set of general principles and Right is underpinned by the fact that liberalism often uses a dichotomy—not a trichotomy—to rebut conservatism. Here we refer to the dichotomy between private and public morality. Roughly speaking, private morality refers to that part of morality where the individual is what we would like to call “morally independent”. What we mean by this is that other persons or intermediate institutions, such as the state, have no morally legitimate status to claim a role with regard to the determination of moral rules or principles as imperative, with regard to their enforcement or with regard to the moral judgments that a person makes in relation to these rules or principles. Independence does not necessarily mean that a person can go about behaving “as she desires or pleases” (see Atterton 2007). The person may be constrained by a duty that follows from her own conception of the good or the conception of a religious or cultural community to which she belongs. Independence primarily means that her fellow citizens or the general public have no legitimate moral status to interfere with her considerations or actions. So if, for example, “the duty to help others” belongs to private morality, the question of whether a person must acknowledge this substantive duty as well as its interpretation in specific situations is beyond the grasp of others. Public morality is commonly conceived of within liberalism as that dimension of morality for which it holds that the state (i.e. the government) has a claim to a legitimate role with regard to the determination of a moral principle or rule, its enforcement or with regard to the judgments made in concrete situations. This dimension of morality may therefore be enforced, since that is the state’s principal way of interacting. Notably, there is no room for a separate category of general principles or rules in this scheme. A moral statement either belongs to the private sphere and is therefore part of the individual pursuit of virtue or it is part of the public sphere and therefore belongs to Right. Consequently we must conclude that liberalism does not commonly—or necessarily—accommodate a quadripartite division of business responsibilities. According to its tenets, “Right” and the full set of general principles cannot be separated.

It follows from this brief analysis of liberalism that the quadripartite division of business responsibilities only makes sense if we either reject the dichotomy between public and private morality altogether or reconceptualize its meaning. Partly in anticipation of arguments to come, we will here clarify our stance in this matter. We believe the distinction between private and public morality to be vital to any understanding of liberalism. This requires that we re-define the concept of public morality, in such a way that it can encompass both Right and the full set of general principles. As we see it, public morality is the dimension of morality for which it holds that the state or others have some basis for claiming a legitimate role with regard to the determination of a moral principle or rule, its enforcement or with regard to the judgments made in concrete situations. Consequently, in matters of public morality, individuals are not independent in the sense described above. If the duty to help others is part of public morality, at the very least a person must accept that she does not have a moral right to ignore others completely. Others have some role to play. Surely, this role will differ in accordance with the various aspects of public morality. The position of others for example greatly differs with regard to Right and the full set of general principles. The exact nature of these differences will be taken up later in our discussion of Neo-Kantian political ethics.

• 2. An interesting aspect of Kant’s liberalism is that he construes a radical distinction between Right and morality, and thus between Right on the one hand and the individual pursuit of virtue and the full set of general principles on the other. In Kant’s thinking, morality is always about self-governance (Kant 1797: 145, 6–380). Morality finds its ground and justification in the reasonable acknowledgement of moral freedom. Reasonable and morally free beings act out of duty (i.e. in accordance with the laws of morality) because the laws of morality strike them as an inescapable fact of reason. But since humans are also natural beings, obeying the laws of morality is far from easy. Still, as rational beings, we are willing to try. Morality in all its forms is fundamentally aspirational.

Right is not about aspiration but about safeguarding certain conditions.¹⁰ Consequently, Kant does not derive the legitimacy of Right from autonomy as a moral (and reasonable) end, as other political theories sometimes do. Right—including its enforceability—follows from the plain fact that freedom (including moral freedom) is only possible in a state in which “the hindering of the hindrance of freedom” is realized. Rational men therefore acknowledge the binding nature of the rulings of the state because of their necessity (Höffe 1992). Thus, even if the rules of state sometimes overlap with moral rules, it is essential to stress that the rules of Right are not moral rules, nor are the claims of the state moral claims (Pippin 2006: 422). Morality and Right always differ with regard to the kind of obligation they involve (Kant 1997: 22).

Having introduced this new distinction, the question is whether Kant is able to provide a framework for a quadripartite division of business responsibilities. In order to do so, he must have the conceptual apparatus to distinguish between Right, the full set of general principles and the individual pursuit of virtue. At this stage in the argument, the prospects look encouraging. Kant follows liberalism in distinguishing Right from the individual pursuit of virtue. In addition, he enforces a radical split between Right and all forms of morality and thus between Right and the full set of general principles. As previously stated, in Kant’s view a society purely based on Right may be possible but it is neither rational nor aspired by humans. Such a society amounts to nothing more than a civitas: a multitude of human beings brought together as citizens under Right. This is a minimally ordered free society in which self-governance is not yet or not necessarily realized. Humans as rational beings envisage another way of living together. We may call this a “kingdom of ends”. In such a society, human moral life (autonomy) is realized and all humans fulfil their moral duties (Kant 1785; see Herman 1997).

Nevertheless, a problem remains as regards Kant providing a shell for a quadripartite division of business responsibilities. The ability to function as a shell for a quadripartite division of business responsibilities requires that Kant can somehow distinguish between the individual pursuit of virtue and the full set of general principles, such that a person remains independent with regard to the pursuit of individual virtue but not fully independent with regard to the full set of general principles. After all, these principles have been defined as “socially acknowledged”. It is not entirely clear whether Kant does this. The fundamental distinction that runs through The Metaphysics of Morals (1797)—Kant’s work on the first principles of substantive morality—is between the sphere of Right and the sphere of morality. Right and morality are distinguished in many ways. Their separate justifying ground is only one dimension. For instance, Kant also distinguishes Right and morality in terms of the typical form that substantive duties take within each sphere. However, the central issue in this context is the lack of clarity as to whether Kant differentiates within the sphere of virtue, such that the distinction between the individual pursuit of virtue and the full set of general principles acquires meaning. On the one hand, this clearly does not seem to be the case. Kant emphasizes that any person is independent in relation to virtue.

That ethics contains duties that one cannot be constrained by others…().. to fulfil follows merely from its being a doctrine of ends, since coercion to ends.().. is self-contradictory.…()…To have an end that I have not myself made an end is self-contradictory, an act of freedom that is yet not free. (1797: 146–147; 6–381; see also 149; 6–385 or 165–166; 6–407)

Since morality is about self-governance, it can therefore only relate to duties about which I can and must decide voluntarily. On the other hand, however, there seem to be signs that Kant does tend towards making the distinction. In relation to the duty to help, Marcia Baron (1995: 163) for example notes that “Kant does not ever say that some acts of helping others are mandatory, but it is plausible to suppose that he would grant that”. What is more, Kant does say that some duties of virtue (respect in particular) are not meritorious but owed to others (Kant 1797: 6–448).

Our cautious conclusion must therefore be that Kant may possibly provide a proper framework for a quadripartite conceptualization of business responsibilities. But we cannot be sure. He is ambivalent when it comes to making a fundamental distinction between the individual pursuit of virtue and the general set of rules and principles requisite for behaviour.

• 3. The proposition that we are defending in relation to the contemporary Neo-Kantians is that they position Kantian thinking more clearly in this respect. Part of their work seems geared to showing that we can make a meaningful distinction within the sphere of the duties of virtue. This distinction coincides with our distinction between the individual pursuit of virtue and the full set of principles. This is not to say that making such a distinction is an explicit Neo-Kantian project. Still, it can be maintained that, implicitly at least, Neo-Kantians are engaged in carving out a sphere within the duties of virtue in which individuals are not completely independent; others may have some moral status regarding the conduct of the individual.

Reath (1997a; see also 1997b) constitutes an interesting example. He argues that Kant does not adequately distinguish between two roles that are both important in analysing duty: the role of the legislator and the role of “source position” or “the source of the reasons for action”. Reath holds that Kant mistakenly thought that if a person herself formally constitutes the authority that demands dutiful action, she must also necessarily be the only one with a say in determining the material or substantive principles by which she ought to abide. Reath argues that this need not be the case. There is no contradiction in maintaining that the ultimate reason why humans accept the full set of general principles is grounded in the aspirational aspect inherent in morality and at the same time maintaining that the substance of what humans ought to accept in that regard, must have a social dimension. For once, there is a distinction between the formal and the material that Kant overlooks. Formally speaking, it is true to say that the individual must independently decide whether to acknowledge the full set of general rules, otherwise morality would no longer be about self-governance. But materially speaking, reasonable people will find it reasonable to determine these duties within a context of social interaction. Stated more emphatically, it is precisely because morality is grounded in reasonableness, that people would find it reasonable to determine these rules socially.

The same line of reasoning can be found in Herman’s work (1997). In discussing the Kingdom of Ends, she rejects the traditional view of Kantian moral man as a lonely person, alienated from others in a pre-political world, only able to connect to others by self-governance. According to Herman, Kantian moral man is deeply social. His basic rational norms of self-legislation are social norms.

We believe that the work of Scanlon on “what we owe to each other” can also be interpreted in this context. When Scanlon (1998: 6) introduces the category of “what we owe to each other”, he explicitly acknowledges that he is breaking through the conventional categorization of morality by stating that the morality of what we owe to each other “is most discussed in latter day contemporary moral philosophy. (However) it is not clear that this domain has a name”. What is more, he also makes it plain that what we owe to each other does not coincide with Right; indeed he explicitly distinguishes one from the other (Scanlon 1998: 4–6). Scanlon also clearly distinguishes this category from the individual pursuit of virtue, explicitly stating that morality is broader than “what we owe to each other” (1998: 171–177).

It therefore seems that the Neo-Kantians are able to construe the fundamental distinctions between (a) Right, (b) the pursuit of virtue by the individual and (c) the full set of general principles. They are able to give an account of these three categories, in which each has an independent meaning. Right concerns the possibility of freedom and is grounded in necessity. Without Right, no form of civil life is possible. The pursuit of virtue by the individual has to be distinguished from Right in two fundamental respects. It is grounded in our rational pursuit of autonomy and with regard to its substantial determination each individual is “independent”. It concerns matters that relate to a person’s own perception of the good life and thus her private morality. The full set of general principles also belongs to morality and is thus grounded in autonomy and aspiration. But at the same time its substance is socially determined. It concerns matters which reasonable people will regard as belonging to public morality, in our more broadly conceived sense.

The Neo-Kantian clarification of the distinctions between the three categories raises a new issue. We must determine the standing of others in relation to the full set of general principles. We know that an individual is independent in relation to the individual pursuit of virtue and that duties of Right may be enforced (by the state). But what is the position of others in relation to the full set of principles? What is the practical meaning of others having a claim?

We will conclude this section by tentatively setting out to clarify the minimum position of others in relation to the full set of general principles. Formulated in negative terms, it is clear that the practical consequence cannot involve others being able to force an actor to do something. After all, the full set of general principles does not belong to Right. In order to produce a more positive account, it is useful to recall that in the Kantian framework all moral duties are “duties of virtue”. Duties of Right and duties of virtue differ as to their structure. Duties of Right are rules that always trump other considerations; a person must always conform her actions to these rules. By contrast, duties of virtue relate to principles that “allow for exception” (Kant 1797: 153–6/390). Given our goal, it is interesting to investigate the exact nature of this “allowance”. It does not mean that a person may sometimes ignore her moral duties. The principles should always be taken into account in one’s decision-making processes. The fact that the duties of virtue “allow for exception” means that one can have good reasons to decide to act on other principles which are also relevant to the situation (see Herman 2001: 214). This makes the moral duties discursive in a quite literal sense. Even if they are dealt with in isolation, they imply narratives and justifications.¹¹

This insight provides opportunities to ascertain what position others may have in relation to the full set of general principles. It suggests that others at least may demand that a person should not deal with matters concerning the full set of general principles in isolation. Discourse is a social event. Thus a person can be asked to inform herself about other views on these matters and reason with at least some others as to the reasonableness of her views (and those of others). We also regard this as being consistent with the views of the Neo-Kantians. Scanlon (1998) for example refers to “what we owe to each other” as “a system of deliberation”.

Deliberation, discourse and the pursuit of a reasonable consensus all sound as if they are very much in the spirit of liberalism and liberal democracy (see for example Habermas 1981; Rawls 1993). Nevertheless, we must proceed carefully here. In this paper we have redefined—and essentially broadened—the sphere of public morality. Public morality has become meaningful outside the sphere of the law. Consequently, there is a certain danger that our attempt to make sense of CSR may in fact endanger the individual freedom from social control that liberalism has fought so hard for in the last few centuries. Our suggestion about the minimum position of others should therefore be regarded as nothing more than a first proposal, one which undoubtedly stands in need of further reflection. Still, we would like to point out that we have tried to accommodate the liberal fear of an intrusion into the sphere of morality that must remain “private” by formulating our principle in the broadest of terms. For example, we have made no attempt to identify “the others”. They might simply be a person’s spouse, friends or colleagues. We are certainly not implying that a person be required to defend her views in some public setting. What is more, by saying that one must reason with these others, we are not suggesting that a person can only end the discussion when a consensus is reached. (In spite of all these precautions, we still think that our conclusion is rather remarkable, especially in a business context where “moral muteness” determines the order of the day (Bird and Waters 1989)).¹²

5 Carroll Placed within the Neo-Kantian Context

Neo-Kantians distinguish between Right, the individual pursuit of virtue and the full set of general principles. It follows that Neo-Kantian thinking provides a good framework for a quadripartite division of business responsibilities. Carroll’s “legal responsibilities” can correspond with Right, his “ethical responsibilities” can correspond with “the full set of general principles” and his “discretionary responsibilities” can be related to the individual pursuit of virtue. Interestingly, we can now also give a positive account of what the ethical responsibilities are about. Actors owe these to each other, but they should not be enforced because they are grounded in morality. Yet others do have some claim with regard to them.

The Neo-Kantian framework will only be available to Carroll if certain changes can be made in his thinking on the division of business responsibilities. Carroll’s view as to the relation between law and morality for example requires radical modification. In the Neo-Kantian scheme, law is not embodied morality. Since the primary focus of this paper is on finding a framework that can justify and make intelligible the quadripartite division of business responsibilities, it is not our task to give an exhaustive account of all the changes needed. Instead, we will concentrate on the changes needed in relation to the substance of the duties Carroll refers to as typically “ethical”, since there seems to be a major problem in relating Carroll’s typical ethical duties to the category of the full set of general principles.

Carroll often includes three substantive principles in the core of his category of ethical responsibilities: “act fair”, “be just” and “do no harm”. The problem with these three principles is that “justice” and “the harm principle” clearly have to be included in Right, both within Neo-Kantian thinking and the broader context of liberalism. However, we think that conciliation is possible with regard to this specific difference. It should be remembered that Carroll is not construing a full-fledged political ethical theory. He is trying to conceptualize CSR from the business ethics perspective of individual market actors. This means that he is taking the perspective of those who are confronted with particular moral issues in the market. Now, any existing society will, to some degree at least, be confronted with a discrepancy between Right as it ought to be institutionalized in a perfect world and Right as it actually works out in imperfect circumstances (failing institutions, difficult socio-historical circumstances etc.). Since these differences are an inescapable fact of life, it seems reasonable to argue that reasonable people determined to live under reasonable rules will acknowledge the principle that each person must make an effort to further justice and undo wrong within the bounds of her institutional role and historical circumstances as a primary duty of virtue. If this is reasonable, then it makes perfect sense for Carroll to put the duties to “be just”, “act fair” and “do no harm” at the core of his ethical responsibilities. As moral responsibilities we must interpret these duties in an aspirational sense (i.e. to further justice).

6 Conclusions and Epilogue

In this paper, we have argued that definitions matter. This compelled us to acknowledge that we cannot be indifferent to the fact that some authors use a tripartite division of business responsibilities as a basis to define CSR, while others use a quadripartite division. We went on to show that any conceptualization of CSR must be justified within political theory. Subsequently we argued that a quadripartite conceptualization of business responsibilities can be justified by embedding it in Neo-Kantian thinking, a theory which can consistently make a fundamental distinction between Right, the full set of general principles and the individual pursuit of virtue. Neo-Kantian theory therefore makes it possible to understand and justify the distinction between legal responsibilities, ethical responsibilities and discretionary responsibilities.

It follows that if Neo-Kantian liberal democratic theories were the only political theory around, we would now be able to argue that a conceptualization of CSR based on a quadripartite division of business responsibilities makes much more sense than any definition based on a tripartite division of business responsibilities. Any definition based on the latter would confuse one relationship in which Neo-Kantian thinking makes a key distinction.

However, we live in a pluralist society. Neo-Kantian thinking is just one of many political theories. Even within liberalism, a strong opponent has now emerged in the form of libertarian thinking (Nozick 1974; Heyd 1982; see also Kymlicka 1990). If we were to try to construe a definition of CSR within libertarian thinking, it would clearly have to be based on the tripartite conceptualization of business responsibilities, since libertarian thinking assumes that all moral duties are enforceable and that there is no such thing as public morality beyond the sphere of Right. This prompts us to consider whether we have accomplished enough in this paper. We think we have. We have achieved our methodological goal by showing that discussions on the conceptualization of CSR must be settled in the context of political theory. We also have grounded the quadripartite division of business responsibilities in contemporary Neo-Kantian political theory. In doing so, we have gone beyond a demonstration of how this quadripartite division of business responsibilities can be justified and made intelligible. At least for those who hold CSR to be a moral duty, we have shown the relevance of some of the characteristics peculiar to Neo-Kantian thinking, in particular the effort to expand public morality—in some form—beyond the sphere of Right.

Of course, the fact remains that this paper is only of preparatory value to those interested in answering the question of whether CSR should be conceived of as a moral duty. In the present context, we can only accommodate these readers by sketching the outline of one important argument in favour of a quadripartite understanding of business responsibilities, as clarified by Neo-Kantian thinking. This argument relates to the cultural understanding of CSR. Culturally speaking, CSR has a strong connotation of being worthy of moral esteem: companies involved in CSR are regarded as doing good. This is neatly illustrated by the title of one of Frederick’s latest books (2006), a survey of the history of CSR since the 1950s: ‘Corporations be good!’. It is also illustrated by Friedman’s position (1970). Despite all his criticism of CSR as an institutional phenomenon, Friedman still expresses respect for individual companies that genuinely display CSR. As he sees it, the management of such a company is fooled by the wrong discourse but still has its heart in the right place.

We hold that libertarian thinking cannot make sense of this cultural aspect of CSR. A libertarian framework forces us to explain the moral worthiness of CSR as originating in its voluntary nature: CSR is of value because the market actors who endorse it voluntarily perform publicly appraised acts. This position does not seem to be very convincing, given the fact that self-interest counts as a proper motive in this conceptualization. Why place moral value on an act performed out of self-interest in the market, even if its consequences are beneficial to society? We might just as well say that these companies are lucky, in the sense that doing the right thing is so easy for them given the fortunate social and historical circumstances in which they find themselves. They enjoy the privilege of being able to combine self-interest with meeting certain social standards.

Neo-Kantian political theory can make sense of this cultural characteristic. CSR is valued and deemed worthy because an actor who endorses CSR has accepted moral duty as an independent ground of action in a given case. After all, ethical responsibilities refer to duties of virtue and the possible reasons for an actor to accept these can only be internal.¹³