Causation and Responsibility in Criminal Law: Excuses that Work and Those that Don’t
About the Author
Samantha Love read Law at Merton College, Oxford, and is currently following the BCL course.
Perhaps my favourite phrase to get me to study is “the harder you work, the luckier you get.”
It’s a good way to remind yourself that whilst some people may seem to have a charmed life, it is more often the fact that they have worked hard to achieve what they have done. And yet, when somebody is lucky in an exam or other important milestone, despite not having worked enough or having the right attitude, they will still be congratulated on their achievement. Equally we might recognise that someone has done a lot of preparation for an exam, interview or even a race and credit them with that effort, but if something outside their control means they do not succeed or win (difficult questions, late bus, a sprained ankle), then the fact that the failure was due to bad luck doesn’t change the outcome. Actions and outcomes matter to society.
This idea is Hart and Honoré’s (1), and was developed to explain why the criminal law does not just punish on the basis of outcomes when they weren’t intended, or just on the basis of bad thoughts when someone intends to perform a criminal act but never achieves it. It is not immediately obvious that you should be punished for an outcome which is disproportionate to what you intended – if, for instance, you intended to harm someone severely but not kill them, but then a doctor’s negligence results in the victim’s death. And yet we do have such doctrines in criminal law, and there are many complex rules on causation which attempt to deal with the question of when a defendant should be responsible for things that happen after they commit a crime – or, indeed, things that don’t happen.
What are we punishing? Outcomes and intentions
Honoré argues that the actual consequences of our actions matter because it is part of our humanity that we have a history and character, which is given to us based on our past activities and their consequences. It is not a system we can opt out of; it is part of being human that we have such identities, which are based on what we do in life. His example of two footballers who need to score in a match is useful for showing this; one makes a bad shot but the wind sends the ball into the net, whereas the other makes a good shot but the wind catches the ball and he does not score. We will recognise the skill in the second footballer’s shot but the first will be the one who has scored a goal.
John Gardner also shows the importance of the difference between outcomes and actions, with his discussion of ‘action reasons’ and ‘outcome reasons’ (2). Lying to a friend is wrong whether or not your friend finds out about the lie, but there is the additional reason for not lying that they would be upset if they discovered that they had lied to you. The ‘outcome reason’ of the friend finding out gives another reason not to lie; that your friendship might end. It is the sort of reasoning which will be familiar to anyone who has disobeyed their parents; those situations where your parents will almost certainly find out are the ones where seems more sensible for you to obey them!
So it’s clear that the outcomes of our actions are considered important. Nonetheless, we will all perform actions that will alter the course of events in one way or another although we could not say that we are in any meaningful way ‘responsible’ for an outcome. My grandmother is not responsible for my exam grades, even though without her I would not have been born – and even if she had helped me with my revision, I was the one who took the paper. Unless she set the paper and told me the right answers, I am the one who put my own thoughts and knowledge down, on which my grade is based. The law uses the causation doctrine to work out which of all the things that come together to cause an outcome ought to be considered significant.
It is worth noting, however, that as well as the causation doctrine there is the need for the mens rea, the required intention, for a certain crime to be present. The question will often be about the degree of harm in the case of freak accidents or interventions by others; the absence of the intention to commit a more serious offence puts a cap on the maximum sentence that can be passed.. For instance, the intention criterion for murder is intent to kill or cause grievous bodily harm. If somebody dies after I push them over, then my sentence will at worst be limited to reckless (constructive) manslaughter because I did not intend to cause death or serious harm. This is an extra way of ensuring that we apportion responsibility correctly; as we saw above with the examination example, outcomes are important but so are intentions and actions. Whilst manslaughter can carry a long sentence, there is a big moral difference between the labels of murder and manslaughter, and more scope to reduce the sentence for manslaughter.
The causation doctrine
The causation doctrine can be boiled down to the question of whether the defendant’s illegal action was an ‘operating and substantial cause’ of the harm which resulted – as set down in Cheshire (3). It would be a mistake to think that this explains it on its own though! This is a legal question rather than a factual one, and depends on common sense and some moral reasoning rather than asking technical medical questions.
We start with the ‘but for’ question – ‘but for’ the defendant’s action, would the harm have occurred? John Herring uses the example of Albert poisoning Victoria, but Victoria dying of an unrelated heart attack before the poison takes effect (4). To put events the other way round however, does make a difference. Shortening the life of someone with a terminal illness is causing their death, because without the illegal conduct they would not have died at that time and in those circumstances.
The ‘but for’ doctrine will however still involve a lot of potential causes – like my grandmother and my exam grades. So we also ask about legal causation: whether the defendant’s action (5) was an ‘operating and substantial cause’ of the harm (6). This is most significant where the action or inaction of another person, or the victim themselves, changes the normal course of events. This is known as a novus actus interveniens – a new and intervening act. Professors Hart and Honoré developed it, using the distinction between those circumstances that are part of the factual background (conditions) and those that are causes. They point out that in order to start a fire you need a dropped match, oxygen and combustible material, but we would only consider the first to be a ’cause’ of that fire. Conditions (oxygen, flammable materials) are normal whereas causes (the dropped match) are abnormal, and in their view only abnormal things can be causes. The question is, of course, what is abnormal?
Hart and Honoré emphasise that only a “free, voluntary and informed act” of a third party can be abnormal and break the chain of causation. A few examples will demonstrate:
– If a person is not acting voluntarily their action will not be a novus actus interveniens, such as falling into another person when pushed and injuring them as well (Herring).
– In Pagett (7) the defendant was using a girl as a human shield against police officers and she was shot dead by a police officer who was aiming for the defendant. The police officer was acting to enforce the law and so his action was not a novus actus interveniens. Hart and Honoré would say that such an obligation meant that the officer was not acting voluntarily, an idea that has been criticised because someone owing a legal duty still makes a choice as to whether to obey that duty (8). What do you think? Is this a public policy exception or does a police officer really not have a choice?
– In Michael (9) a mother gave her baby son’s nurse some laudanum, telling her it was medicine for her son and intending the nurse to poison and kill him. The nurse put it on a shelf and forgot about it, but her child gave it to the baby and he died. The nurse had no knowledge of the bottle having poison in and the child was too young to be capable of criminal conduct, so the mother was found guilty of murder through the use of innocent agents.
The act must also be enough to render the defendant’s act no longer a substantial and operating cause. This is particularly important in criminal cases that involve the subequent complication of medical negligence, where the courts are reluctant to let defendants complain that their victim would have survived if they had received proper medical care. A rare case is R v Jordan (10), where the victim’s wound had almost healed when doctors gave him a drug they knew he was severely allergic to and he died.
As for actions of the victim, two important cases are Roberts (11) and Blaue (12). In Roberts, the victim had been given a lift by the defendant but jumped out of the car when he tried to molest her, and injured herself. Jumping out the car was found to be a reasonably foreseeable reaction and so the defendant was responsible for her injuries – though Stephenson LJ did say that if the victim did something ‘daft’ and unforeseeable then the defendant would not be responsible. In Blaue the defendant stabbed the victim, who would have survived had she not refused a blood transfusion because she was a Jehovah’s Witness. The court said that the defendant had to take his victim as he found her, and so he could not say that her death was her own fault.
There are a few ways to think about these two cases. Is this about foreseeability? Should we say it was foreseeable that the victim in Blaue would refuse a transfusion? Or perhaps that it’s about voluntary acts; the defendant in Roberts acted without thinking because she was under attack, and in Blaue the defendant’s religious views meant that she could not voluntarily choose other than to refuse treatment? Bruder contests however whether our religious views make our choices completely involuntary – you can after all change religion or break religious rules as a matter of actual fact, even if you feel that you can’t. What if somebody’s religious views meant that they had to maim women? Would we say that their acts are involuntary? Or perhaps these are just two different circumstances that we are protecting for different reasons – one because the defendant should have foreseen that his victim would jump out of the car, and one because we do not want to blame a victim for her religious views?
These two cases demonstrate how tricky the law on causation can be to understand. The general principles hide a lot of difficult questions, and the court will always try to do justice even if the reasoning may sometimes seem out of line with previous cases. Academics worry about the reasons for the criminal law and how this influences rules on causation, but it’s difficult to know how far all of this follows a single grand plan. After all, as Hart and Honoré say, our ideas of responsibility are part of being human and part of a society. So it makes sense that our rules on causation will follow our general understanding of causation, even if that doesn’t fit a perfect set of legal principles.
(1) Hart and Honoré, Causation in the Law (2nd edn, Oxford University Press 1985)
(2) J Gardner, ‘Responsibility and Luck’  104 Law Quarterly Review
(3) R v Cheshire  3 All ER 670
(4) J Herring, Criminal Law, (7th edn, Palgrave MacMillan 2011)
(5) or inaction, in certain limited situations
(6) R v Cheshire, ibid
(7) Pagett (1983) 76 Cr App R 279 (CA)
(8) Bruder, ‘Owning Outcomes: On Intervening Causes, Thin Skulls, and Fault-undifferentiated Crimes’ (1998) 11 Can J L Juris 90
(9) Michael (1840) 173 ER 867
(10) R v Jordan (1956) Cr App R 152 (CA)
(11) R v Roberts
(12) R v Blaue