Most of the events that are discussed on the Rethinking SLIC blog through the lens of secondary liability are exceptional. They are exceptional, when one considers that most instances of human cooperation and aid are perfectly unobjectionable, both legally and morally. Financial services, for instance, are indispensable to the modern world. Accordingly, earlier blog posts from the project considering how such services have contributed to serious wrongs concern exceptions. Examples are Jindan-Karena Mann’s blog concerning BNP Paribas in Sudan and Joëlle Trampert’s blog on FMO’s involvement in the construction of a dam on the Qualcarque river.
The general rule in relation to which scenarios of secondary liability form exceptions is often construed in consequentialist terms. The idea is then that, in general, secondary actors contribute to good things. This consequentialist rationale supports having relaxed rules of secondary liability, so that beneficial cooperation is not discouraged. Georg Nolte and Helmut Philipp Aust for instance write in this vein that
[w]hile [...] a strict rule on the responsibility for ‘aid and assistance’ at first sight appears to be beneficial for the international rule of law—as it would claim to force States to be their ‘brother’s keeper’ and to steer far away from the risk of being implied in illegal activity—it would at the same time discourage many typical and usually beneficial forms of international co-operation (p. 12).
The question of when exceptions should be made to this rule, that is, when secondary actors should face liability, is also understood in consequentialist terms. That is to say, when secondary actors contribute to serious wrongs, the law on secondary liability should come in to hopefully prevent such contributions from being made. To give an example, when the financial services provided by a company make part of a causal chain leading up to serious human rights violations, consequentialism supports attaching liability to the provision of the services, instead of supporting the provision of the services itself. Consequentialism can thus provide both arguments for and against having an expansive regime of secondary liability.
In this blog, we consider a different rationale that can provide normative guidance in rethinking the law of secondary liability: respect for autonomy. This is an idea we gather from deontological thought. It is our impression that respect for autonomy has not been granted much attention yet in deontological theories of secondary liability, which instead focus on asking why and when secondary actors deserve punishment or should be made to rectify wrongs. To illustrate what we mean by respect for autonomy and how this idea can inform normative arguments about the law on secondary liability, we turn to an infamous passage from Immanuel Kant’s On a Supposed Right to lie from Philanthropy (1797): The murderer at the door.
II. The murderer at the door
What should you tell a murderer, Kant asks, when he knocks on your door and asks you whether your friend, who is hiding upstairs, is around? Kant’s answer is that you should, if pressed for an answer, not lie. Kant goes further than this, writing that “if you had lied and said that he [your friend] is not at home, and he has actually gone out (though you are not aware of it), so that the murderer encounters him while going away and perpetrates his deed on him, then you can by right be prosecuted as the author of his death” (8:427). This passage from Kant’s writing is the subject of much controversy. Many commentators for instance maintain that Kant applied his own theory incorrectly to the -fictional- facts. We do not want to take sides in any interpretative debate here. Instead, it is interesting to note that the scenario of the murderer at the door may involve secondary liability: If the murderer succeeds in his plans, the person who answered the murderer’s question may have contributed to this, either by telling the murderer where to find his friend or, in the scenario Kant also discusses, by lying to the murderer who then by coincidence runs into the friend regardless.
Kant makes two related moral claims about his fictional story. The first is that you are not responsible for the consequences of telling the murderer the truth. The second is that you are, in sharp contrast, responsible for any consequences that may follow from lying about your friend’s whereabouts. We use the first idea in the remainder of our argument, leaving the second -strange- claim aside. What underlies both claims, however, is Kant’s conception of persons as free and rational agents. When Kant speaks of freedom, he understands this as the ability to overcome one’s urges. As rational beings we have the capacity to know what is right and wrong and to act on this (or so we must believe, at least). Rationality allows us to free ourselves from our impulses by doing the right thing. According to Kant, this quality separates humans from inanimate objects and other animals, who cannot resist their urges when reason would compel us, if we were in a similar situation, to act otherwise. Instead, what happens to objects and non-human animals is merely the result of causal laws. Accordingly, Kant makes a sharp distinction between things that behave fully according to the laws of nature and persons, who operate (partly) outside of this causal realm.
This distinction has important implications for questions of responsibility. That is because, in Kant’s framework, responsibility can only reside with persons.1 When only causal objects stand between a person’s action and its eventual consequences, that person can, if other circumstances -the presence of the required mental state, etc.- also allow for it, be held accountable for the consequences they produce. If, on the other hand, one actor relates to another as a secondary actor, the principal actor is in the first instance solely responsible for what they themselves ends up doing. Christine Korsgaard makes this point in discussing the murderer at the door as follows: “In a Kantian theory our responsibility has definite boundaries: each person as a first cause exerts some influence on what happens, and it is your part that is up to you. If you make a straightforward appeal to the reason of another person, your responsibility ends there and the other's responsibility begins.” (p. 337).
Positively, the implication of this argument is that persons should not approach others as if they are causal objects. It is respect for the autonomy of others that leads Kant to the idea that we should not, through deception, try to make other persons act differently. Negatively, it means that what someone ends up doing with the help of another should generally not be held against that other. Accordingly, in Kant’s story, the wrong you can commit is the relatively minor -we would say- wrong that is lying. The bigger wrong of murder looming in the background is not for you to worry about.
III. Respect for autonomy as a criterion for rules of secondary liability
Now clearly, neither Kant’s view that lying makes you responsible for all consequences that follow from it nor his claim that telling the truth can never be wrong are reflected in the law. The Dutch Criminal Code for instance explicitly mentions intentionally providing information as a ground for accomplice liability (art. 48-2), rendering the argument that telling the murderer the truth is criminal a simple one. We also do not want to argue that this conclusion is wrong or that in light of Kant’s argument some more general large overhaul of the law should take place. The law on secondary liability is not strictly Kantian –or consequentialist, for that matter- but reflects and should reflect elements of both theories as well as other considerations.
Our more modest claim is instead that, next to other considerations, there should also be room -more room- for the Kantian concern of respect for autonomy in normative discussions about secondary liability. It is our impression that most normative arguments made in discussions surrounding secondary liability are, often implicitly, consequentialist. The argument that stricter rules of secondary liability could prevent serious wrongs is for instance consequentialist. And so is the opposite argument that a strict regime can also be too strict, when cooperation that would lead to good things is discouraged. But also arguments in discussions about more specific regimes can often be understood in consequentialist terms. An example is the argument sometimes brought against due diligence regimes that such regimes may lead to a mere ticking of boxes that absolves businesses from liability while failing to prevent the wrongs it is meant to prevent. Next to arguments of this kind, which are important, we argue that a different sort of argument that does not look at the consequences of the acts of secondary actors but instead asks whether what eventually happens is for the secondary actor to worry about in the first place, should not be overlooked.
This leads us to two takeaways. Firstly, recognizing this kind of argument may allow us to understand existing law better, in so far as the law reflects concerns related to respect for autonomy. Secondly, adding respect for autonomy to our normative toolbox may at times also lead us to different conclusions.
For both of these points, it is important to note that considerations based on respect for autonomy and consequentialist considerations do not always coincide. That is to say, the arguments that follow from these considerations in favour of or against holding a secondary actor liable are different and can support opposing conclusions. The above discussion has also made clear that arguments grounded in respect for autonomy serve to support the conclusion that secondary actors should escape liability. But, when concern for autonomy is not a concern, arguments in favour of liability can be accorded more weight. Consequentialist, retributivist, and other arguments for far-reaching liability then find their full force. Accordingly, even though respect for autonomy can only pose an obstacle towards holding secondary actors responsible, this does not imply that we argue for less secondary liability.
In the following section, we aim to illustrate how respect for autonomy relates to questions of secondary liability by reflecting on a number of central notions and types of situations relating to secondary liability from the perspective of respect for autonomy.
IV. Respect for autonomy and concepts of secondary liability
‘Autonomy’ combines the Greek words ‘autos’ (αὐτός) and ‘nomos’ (νόμος) which can be translated as ‘self’ and ‘law’ or ‘rule’. Autonomy, then, is self-rule. Respect for autonomy accordingly gives a reason not to (legally require secondary actors to) interfere with the projects of persons that are, in the relevant respect, really theirs. Turning to scenarios that may involve secondary liability, what we propose is that when a (criminal) project is really the project of multiple actors working together, the idea of respect for autonomy finds no application, because it is contradictory to say that a secondary actor has reason not to interfere with a project that is his own. The less the principal’s enterprise is also that of the secondary actor, the more respect for autonomy becomes a concern.
When can a secondary actor be said to be a co-author of the principal’s project? Both knowledge of the contribution and intent to bring about the crime as well as a larger factual contribution could be seen as supporting this conclusion. Authorship points first, however, to a strong mental element. No matter how large a factual contribution a secondary actor makes, in the absence of intent, it would be strange to regard the actor as a co-author of the crime. In cases in which secondary actors only have knowledge of the crime and their contribution to it, respect for autonomy is not a concern. This is precisely the point of Kant’s parable: seen from the idea of respect for autonomy, the foreseeability of some undesirable consequence is not a sufficient reason to require the helper to act differently, as the responsibility for this consequence lies, prima facie, with the future principal offender.
As an application of the idea of respect for autonomy and its corollary of authorship to the law of secondary liability, consider the situation, also discussed by Nicky Touw in a recent blog post on this website, of multinational corporations (MNCs) that purchase cotton which can be linked to Uyghur forced labour. Should these companies face liability? The rationale of respect for autonomy neatly tracks the conditions under which lawyers would be tempted to reply “yes”. The more involved companies are with the rest of the supply chain, the less respect for autonomy poses a problem to holding them responsible for what occurs in supply chains. It is uncontroversial that consumers should not face liability for standing at the end of blood-ridden supply chains, for instance by wearing cotton with genocidal origins. On the other end of the spectrum, companies which only have shell corporations of their own making standing between them and serious human rights violations are really the authors or co-authors of the wrongs upstream and should accordingly be held legally responsible.
In this blog, we turned to Kant to reflect on current discussions surrounding secondary liability. These discussions largely take place in consequentialist terms. This makes sense, for the need to revisit questions of secondary liability is grounded on the increasing influence that primarily MNCs and states can exercise over others in our current age. Consequentialism provides the language to scrutinize powerful actors and their actions by being focused on the consequences they produce.
Yet it would be wrong to link increasing influence to increasing responsibility simpliciter. Why is it uncontroversial that consumers should not face liability for buying products in the supermarket coming from blood-ridden supply chains? Or why should powerful states and corporations not indeed be required to be their ‘brothers’ keepers’ whenever they could? Adding respect for autonomy to our normative toolbox may help us to understand our shared intuitions when it comes to uncontroversial questions of this sort, and also to possibly reach better conclusions when things are less clear.
1 For our purposes, it is important to note that this should be understood to include entities in which persons work together. Some argument is needed to go from natural persons to corporations and states. For what we aim to say here, it should suffice to say that incorporated agents can do things, that their members’ capacity to reflect on what they do does not disappear when they act together, and that accordingly there is no clear reason why the idea that one should prima facie not be required to interfere with others’ projects does not equally apply when such interference would come from a group of persons acting in close cooperation. The subsequent question of responsibility -whether corporate agents can be responsible and what this would mean and should entail- is a separate question.