1. A Case
Let us start with a case. On Saturday November the 11th 1989, Luis Berrío drove a SUV heading to Tapao, a town in Colombia. The day was sunny and clear. Mr. Berrío was confident about the whether conditions and drove breaking in a slightly the speed limit. He was completely aware of the fact that the 13 year old boy Carlos Jiménez rode a bike along the same road. The boy was not paying attention to the traffic and tried to cross the road without looking to the cars coming. Mr. Berrío reduced the speed of the car and beeped a couple of times but could not avoid running over the boy. The boy died some months after the accident.
2. Strict Liability and the alleged Challenge to the Principle of Control
In many countries strict liability is to be applied in cases of traffic accidents. Liability always presupposes the existence of a harm, an act causing the harm –traditionally understood as a “wrong”– and a causal relation between the latter and the former. However, there are two different systems of liability: fault liability and strict liability, and both of them presuppose an additional element. While the system of fault liability presupposes the existence of a specific mental state in the agent: bad will or fault (negligence is a species of fault), strict liability presupposes that there is a legal rule assessing responsibility to the agent, in spite of the existence or inexistence, the evidence or the lack of evidence regarding any mental state of her. When strict liability is at stake, it is enough for the victim to proof the harm, the act causing the harm and the causal relation between the latter and the former. The agent can only be excused if she proofs that one of these elements is missing. While fault liability holds people responsible for harms attributable to their culpable agency, strict liability holds people responsible for harms attributable to their agency tout court.
The system of fault liability is applied normally in tort’s law cases. The reason is clear: this system is more according to the so called principle of control. The principle of control is an essential principle for the allocation of moral and legal responsibility. Nelkin defines it in the following way:
“We are morally assessable only to the extent that what we are assessed for depends on factors under our control”.
The system of fault liability is more according to this principle because when it is applied, legal responsibility and the consequent liability for the harm is assessed to an agent because of her intentions and actions and not merely because of the outcome. Intentions and actions are (or at least would be more) under the agent’s control. However, the outcome is not always under the agent’s control. Strict liability is a special case of the so called “outcome responsibility” in which the agent’s responsibility is assessed no merely because of her intentions or actions, but because of the outcomes of her intentions and actions, despite that sometimes the outcome is beyond the agent’s control.
Therefore, strict liability –and outcome responsibility in general- allegedly challenges the principle of control. Moreover, As Nelkin explains with respect to moral responsibility, it is intuitively compelling that the following is a corollary of the principle of control:
“Two people ought not to be morally assessed differently if the only other differences between them are due to factors beyond their control”
An application of this corollary to a case would be the following:
“if two drivers have taken all precautions, and are abiding by all the rules of the road, and in one case, a dog runs in front of the car and is killed, and not in the other, then, given that the dog’s running out was not something over which either driver had control, it seems that we are reluctant to blame one driver more than the other”.
However, if (from the moral point of view) outcome responsibility (and in the law: strict liability) is applied, the result won’t be according to this corollary. In fact, the driver who killed the dog would be held moral responsible (and maybe also legally liable). The other driver would neither be held responsible nor liable for any harm (outcome) can be imputed to him.
3. Outcome Responsibility and Identity
Nevertheless, outcome responsibility seems to be crucial to our identity as persons. As Honoré points out:
“If actions and outcomes were not ascribed to us on the basis of our bodily movements and their mental accompaniments, we could have no continuing history or character”.
Thus, outcome responsibility is crucial for identifying us as persons during time; that is to say, as a things extended in space and that persist through time. This is the metaphysical or “reindentification” approach to identity. Outcomes related to our intentions and actions that have been imputed to us, are a basis to determine that we are the same persons along time.
Furthermore, outcome responsibility seems also to be essential for defining identity from a completely different point of view: the characterization approach. This –psychological- point of view defines the identity of a person on the basis of her main traits or characteristics. According to Schechtman, responsibility has a close relationship with this approach to identity. She says that
“the degree to which an action expresses someone’s identity is a central consideration in making moral judgments about it”
and also that
“the degree to which an actions expresses a person’s identity is, moreover, precisely what the characterization question seeks to determine”.
Thus, according to Schechtman, the traits characterizing a person are a relevant element to determine the extent or the degree of responsibility of this person for her actions. She says:
“It is a commonplace that a person can be more or less responsible for a given action that occurs in his history –a person is less responsible for tripping on a rock and knocking someone down inadvertently than he is for willfully pushing someone out of the way when he is under duress, or here is some compelling reason to do so, than in ordinary circumstances”
In addition to this, it seems plausible to think that the outcomes associated with the actions and not merely the intentions and the actions are relevant also factors to characterize the identity of a person. The fact that a person repeatedly performs an action with certain outcome can be associated with a relevant trait characterizing her identity. In this sense, Schechtman says that this “characterization” approach to identity
“asks which actions, experiences, beliefs, values, desires, character traits, and so on [… characteristics] are attributed to a given person”
It is plausible to think that when we attribute to a given person “actions” and “experiences” we do not attribute to her merely primitive actions like moving the finger in order to pull the trigger, but outcomes like killing someone. In addition to this, outcomes are essentially related to the roles that a person plays in society and that are other sources of traits or characteristics of identity.
If this claim is sound –I think that it is-, then outcome responsibility is essential for defining identity from both points of views: the metaphysical point view of the reidentification of the person through time and the psychological point of view of the characterization of the person.
4. The Dilemma
If all what has been said is true, then a dilemma emerges. Either we adhere to the principle of control and maintain that outcome responsibility (and its species: strict liability) cannot be assessed, or we acknowledge that it is justified to asses outcome responsibility (and, therefore, also, strict liability) and we give up the principle of control.
5. Moral Luck
A strategy to accept the second horn of this dilemma is to acknowledge the inevitable existence of moral luck. According to Thomas Nagel, moral luck can be defined in the following way:
“Where a significant aspect of what someone does depends on factors beyond his control, yet we continue to treat him in that respect as an object of moral judgment, it can be called moral luck”.
It is well known that there can be various forms of moral luck. Nagel distinguishes between resultant, circumstantial, constitutive and causal moral luck. Resultant moral luck is about the consequence of actions and situations; circumstantial moral luck depends on the circumstances or the environment surrounding the agent; constitutive moral luck has to be with her personal character and causal moral luck with determinism, that is to say, with the view according to which actions are not the result of free choices by the agent but are determined by external events beyond her control.
From these forms of luck here is only relevant resultant luck, because of its relation with outcome responsibility and therefore with strict liability. Resultant luck concerns the way in which things turn out. In the cases of the drivers, they could have the same intentions and could perform the same primitive actions, but if things turn out differently and one of them runs over a boy but not the other one, and in spite of this fact, we held responsible just one of them, then we have a case of resultant moral luck. If we find that the existence of resultant moral luck is justified and therefore we impute responsibility for outcomes, not just for intentions and actions, and we apply strict liability, then we would agree with the second horn of the dilemma stated above and we would neglect the principle of control.
Now, John Goldberg and Benjamin Zipursky have claimed that there is another form of moral luck that has also a relationship with strict liability. This additional form of moral luck would be the “compliance luck”. According to them, compliance luck is:
“luck affecting one’s ability to meet a relevant norm of conduct” .
The concept of “norm of conduct” is to be interpreted in a broad sense. In this way, compliance luck would be present in cases of negligence in which the agent, because of her fault, fails to fulfill an “objective standard of care” and in cases of strict liability in which the diligence, negligence or fault of the agent does not play any role at all. As an example of the first kind of cases, Goldberg and Zipursky mention the following:
“The classic doctrinal exemplar of compliance luck is the 1837 decision in Vaughan v. Menlove. Menlove stacked hay in a manner that created a risk of spontaneous combustion. The haystack later ignited, causing a fire that damaged the property of his neighbor, Vaughan. Menlove’s lawyer argued that his client had acted in the good faith belief that he was behaving prudently and thus should not be held liable to Vaughan. The court disagreed and famously held that negligence law sets an “objective” standard that measures conduct against the care a person of ordinary prudence would have taken under the circumstances”.
Now, concerning the second kind of cases they mention the following case:
“the tort of trespass to land enjoins each of us to refrain from interfering with property owners’ rights of undisturbed possession. It is possible that a defendant who is acting in an impeccable manner will end up committing a trespass. Suppose D, prior to fencing his yard, consults all relevant records to determine the location of the property line between his property and his neighbor P’s property. Suppose also that D builds his fence strictly in accordance with the information in those records. If it later turns out that the records were, unbeknownst to anyone, erroneous, such that D’s fence is in fact sitting on P’s property, P has a cause of action against D for trespass. As in Vaughan, this is an instance of bad compliance luck. D violated a norm of conduct forbidding one from physically invading another’s property, but D’s violation did not stem from any lack of diligence on D’s part. Insofar as the liability is generated by features of the defendant’s action that would not have been visible or otherwise accessible to a person in the defendant’s circumstances who was conscientiously choosing a path of conduct, liability is turning on bad luck.”.
In this second case, the essential element of strict liability appears, namely, that the agent objectively “violated a norm of conduct” and that this violation as such generates responsibility, without any further consideration regarding the intentions of the agent, her fault or her epistemic situation. However, it is easy to see that “resultant” luck and “compliance” luck are just two faces of the same coin. In cases like this, if moral luck is at stake, the agent does not fulfill the requirements of the norm because her intentions and actions turn to a result or an outcome beyond her control. In this kind of cases, the result or outcome (for instance: not running over pedestrians) is the very same “action” described in the norm that is not fulfilled by the agent (for instance, the driver who abides by all the rules of the road).
Now, this relationship is not present in cases of negligence like Vaughan, where, in my opinion, the lack of fulfillment of the requirements of the “objective standard of care” is independent from the result as such, that is to say the harm. In cases of negligence, it is possible for the agent to violate the “objective standard of care” without generating the bad result or outcome. However, it seems to me that in these cases, if luck plays a role, it is in the form of resultant luck and not in the form of compliance luck. For instance, in the Vaughan case the violation of the objective standard of care is not a matter of luck. What is a matter of luck is the explosion, that is to say, the result or the outcome.
For these reason, I disagree with Goldberg and Zipursky about the relevance of “compliance luck”. It seems to me that in cases of negligence, it is resultant and not compliance luck what is relevant from the point of view of the problem of moral luck and the imputation of moral and legal responsibility. Nowadays, in cases of strict liability, the concept of compliance luck indeed emphasizes an aspect of the role that moral luck plays in the law, namely, the lack of fulfillment of a legal norm. This aspect cannot be captured by the concept of resultant luck as such. However, from this two concepts resultant luck has a priority because the violation of the norm is just the consequence of the production of the outcome that is allegedly beyond the agent’s control. In what follows, he will talk about this sense of resultant luck that is essential for compliance luck in strict liability.
Now, accepting resultant luck in this sense in cases of strict liability generates some problems. First, as Keating points, out such a system of liability would suspect because “of the great burden that it places on our freedom: We may find ourselves liable for devastating injury no matter how carefully we proceed, no matter how irreproachably we conduct ourselves”. For instance, in cases like the described above, if strict liability is applied and resultant luck is admitted, then, it does not matter whether the driver had abided by every rule of the road or whether he has been negligent. The reasoning is just as follows: if the outcome consistent of running over a boy has been produced, he had to be considered liable. Second, the acceptance of this kind of resultant luck changes the original idea of tort’s law as a law of wrongs and makes it become a system of distributive justice. Third, it also makes torts law arbitrary, incoherent and depending on luck. Difference in treatment of similar cases could not be justified. And fourth, it does not seem to offer a full account of strict liability, especially concerning an essential element of this system of liability: the idea that in spite of the fact that the outcome does not depend entirely on intentions and actions of the agent, having these intentions and performing these actions, the agent is aware about the possibility of the outcome and creates the risk of the occurrence of that outcome.
6. Strict liability and factors beyond the agent’s control
Therefore, I would endorse here the following claim: that the awareness of the agent about her own action includes the awareness about the risk of the occurrence of the bad outcome that the agent is creating by her action. This awareness leads us to recognize that in strict liability there are in deed no factors beyond the agent’s control and to deny the existence of moral luck in this field. In other words, I will claim that awareness about the own actions imply awareness about the risks of bad outcomes. If this is the case, then, it is possible that we do not control the bad outcome, but we do control the creation of the risk of a bad outcome. As a result, the existence of moral luck has to be denied, if, in this way, we take risks seriously.
In order to justify my claim, I will begin with an explanation of the concept of self-awareness. One of the most accurate concepts of self-awareness is to be found in David Velleman’s account of action. In this account, one basic element of every action is the awareness of the agent about what she is doing. In Velleman’s opinion, one essential element of the process of doing something is to know what one is doing. In his words:
“The reason why you usually know what you’re doing […] is that you simply don’t do anything unless, first, you have already anticipated doing it next and, second, you have either just become aware of being just about to do it or just started paying attention to what you’re doing”.
He also says that:
“Your desire to know what you’re doing moves you to do whatever you expect to do next and restrains you from doing otherwise. It thereby ensures that whatever you do, you will already have anticipated –which is a means of ensuring that you know what you’re doing”.
Now, Velleman defines self-awareness as the ability:
“to name what you’re doing –to give a prima facie description of your action”
The expression giving a prima facie description of the own action is ambiguous. In a narrow sense, it could mean giving a prima facie description of primitive actions like bodily movements. This sense is not relevant at all for assessing moral and legal responsibility. Relevant is a broad sense, in which to give a prima facie description of the own actions imply describing, as Velleman says, larger actions
“such as walking back downtown, which will occupy more than an instant”
Velleman’s account makes it clear that in this process of being self-aware of our own actions, it is unavoidable to “anticipate” or foresee this kind of larger actions, in order to
“know what sort of action it will turn to be”
It seems to me that at this point it is necessary to make a distinction between actions that cannot produce bad outcomes and generate strict liability of the agent, and actions that can produce this kind of outcomes and generate this kind of liability. Let me use as example of the first kind of action: brushing the teeth and of the second one driving a car. By performing both kinds of actions the agent is self-aware, that is to say, he is able to give a prima facie description of them. In fact, as Velleman also claims, self-awareness is a step in the process of the action that leads the agent to perform the action and to refrain performing another one. It is also true that both kinds of actions can be regulated by “regulative rules”. It is possible to perform both kinds of actions, independently to the existence of moral or legal rules. However, there is an essential difference between them, namely, that moral and legal regulations of the second kind of actions, e.g. driving a car, include descriptions not only of the action as such but also of the risks that the action entails. For this reason, an appropriate prima facie description of the dangerous actions entailing strict liability such as driving a car, shooting a gun, manipulating explosives, etc., entails a prima facie description of the risks that these kinds of action implies. Consequently, it is impossible for an agent to be aware of her performing of the action: e.g. manipulating explosives, without being aware at the same time about the risks this action entails.
In this way, if we consider the case at the beginning of this paper, it is impossible for Mr. Berrío to deny that the description of the action consisting in driving a car (and much more after he saw the boy on the road) entails the description of the risks that this action is creating. This explains also why this kind of actions are severely regulated, why the agent has to be old enough to be aware of the risks and why in most cases it is necessary of obtain a license before performing the actions. As Enoch and Marmor say, by the time of performance of these dangerous activities the agent had have the opportunity to “internalize” the risks.
As a result, the self-awareness of an agent concerning her performance of a dangerous activity like driving a car or manipulating explosives necessarily implies her self-awareness about the risk that her action is creating. In this sense, it is plausible to say that there is a necessary connection between the performing of the action and the creation of the risk. Both are the two faces of the same coin.
8. The Denial of Moral Luck in Strict Liability
Now if the agent is aware of the risks she is creating by performing the action, it is plausible to say that the creation of the risk is not beyond the control of the agent. The creation of the risk is entirely regulated by the principle of control. The agent can be assessed for the creation of the risk.
However, only this circumstance cannot provide a basis for denying the existence of moral luck. It is true that the risk of a bad outcome in under the agent’s control, but not the bad outcome as such. If we take into account only the awareness of the creation of the risk, moral luck apparently cannot be denied. The bad outcome as such seems still to be beyond the agent’s control. However, if we examine things closer the conclusion has to be different.
This is easy to see in the case of Mr. Berrío. By driving he is aware of the action of driving and also of the risk that he is creating when he drives. In fact, as Velleman would claim, this awareness leads him to keep driving and not to perform another different action. According to the description of the case at the beginning of this paper, it is easy to discover that he is also aware about the risks that he is creating. This awareness explains why he slows down when he sees the boy riding the bike and why he beeps a couple of times. If we consider the fact that he was breaking the speed limit, from the moral point of view he could be held morally responsible for that. However, there is something more: he could also be held morally responsible for increasing the risk of achieving a bad outcome. In every action, there is a direct relationship between the increasing of the risk of achieving the outcome and the actual achievement of this outcome. And this conclusion can be generalized to every kind of actions.
For instance, when Velleman discusses in Practical Reflection his account of action, he talks about the action of opening a window. An agent can start opening a window. In a given point, she becomes aware of her action and she can give a description of the action called opening the window. This awareness leads her to keep doing the primitive actions (bodily movements) that are the essential elements of the larger action called opening the window. The more the agent makes these primitive actions, the higher are the risks that she achieves the outcome, namely, to open the window. Now it could happen that the agent becomes aware about the fact that the window is sealed and the glass is very weak and old. In this case, the agent becomes also aware of the following risk: if she keeps trying to open the window there is a high risk that the glass breaks. At this point, she can stop or keep opening the window. We know that she does not want to break the glass but she is aware of the increasing of the risk of breaking the window and she decides to neglect the possibility of having a bad outcome and take this risk.
At this point, due to her awareness about both possibilities, two outcomes can happen: either she opens the window or she breaks the glass. It is my opinion that due to her awareness, no matter what the outcome is, this outcome can be consider the effect of the agent’s action and, therefore, the agent can be held morally responsible for the outcome. The alternative possibility would be saying that the (good or bad) outcome was caused by luck. However, the problem would be that this account can be applied to every possible action and, therefore, the good or bad luck would end being the cause of every good or bad outcome and no space at all would leave for responsibility. The luck would be responsible for everything. This would be a laic but esoteric version of the modern philosophers’ view that attributed to God the responsibility for every good and bad outcome in the World. The World would be absolute determined by the luck.
This same argument can be applied in order to deny the existence of bad luck in cases of strict liability. Strict liability is connected to the performance of actions that can be dangerous for society. These actions are very well regulated in society and the regulation makes it possible for every agent to be aware that the description of the action implies the description of the creation of risks. In this sense, being aware of the action implies being aware of the creation of risks. If this is the case, strict liability leads the agent to take risks seriously, because the bad outcome can be imputed to her. This is a very effective way to protect society against the arbitrary and irresponsible creation of risks. In addition to this, the performance of these kinds of actions generally implies a profit or a benefit for the agent. If this is the case, then it is fare to state the following formula: if the outcome is good then the benefit corresponds to the agent and it the outcome is bad then the agent is liable for the damages he caused. On the contrary, it would be unfair to allow an asymmetry between benefits and damages, that is to say, between good and bad outcomes, and allow the agent to take the benefits but not the responsibility for the possible bad outcomes. It is the agent who damages other people who has to restore them in their goods. It is true that the actions of the agent only partially determine the good or bad outcome. However, the concept of self-awareness shows that the part determined by the action is relevant enough to justify the imputation of moral and legal responsibility to the agent.
Informações Sobre o Autor
Carlos Bernal Pulido
Profesor de Derecho Constitucional y Filosofia del Derecho en la Universidad Externado de Colombia