Why intentions matter to the criminal law

The intentions with which you acted matter to the law in a couple of different respects:

(1) The fact that D had a particular intention in acting as he did can make his conduct unlawful, when if he did not have that intention his actions would have been perfectly lawful. In such a case, we can say that D’s intention was wrong-making.

(2) The fact that D had a particular intention in acting as he did can make his conduct more seriously wrongful than it would have been had he not had that intention. In such a case, we can say that D’s intention was aggravating.

Wrong-making intentions

Consider the following two cases. In both cases, Patient is terminally and is in crippling pain.

Allowed: Dr Good gives Patient a dose of morphine that is sufficient to relieve his pain, realising at the same time that if Patient receives a dose of morphine of that quantity, then Patient’s death is likely to be significantly accelerated. There is no other means of relieving Patient’s pain. Patient subsequently dies from the dose of morphine that he received.

Not Allowed: Dr Bad gives Patient a dose of morphine of exactly the same quantity as that administered by Dr Good. But when Dr Bad administers the dose, he intends that the dose that Patient receives will kill him, thereby ending his torment. Patient subsequently dies from the dose of morphine that he received.

(It should be noted that there is actually no medical evidence that doses of morphine administered in sufficient quantities to relieve pain can actually accelerate people’s deaths; but that does not affect the discussion below.)

The only difference between these two situations is that in Allowed, Dr Good does not intend (in the sense of acting with a particular aim or purpose) to kill Patient, whereas in Not Allowed, Dr Bad intends to kill Patient.

In every other respect, the situations are identical. Both doctors act with the ultimate intention of relieving Patient’s pain (note that you can have more than one intention in a given situation because you can have more than one aim or purpose in acting in a particular way) – it’s just the means by which they intend to relieve Patient’s pain differ. And in terms of the doctor’s physical actions, they are identical: both administer the same dose of morphine, and both bring about Patient’s death at the same time.

So the only difference between the two situations is what is running through the doctors’ heads at the time they inject the morphine into Patient. But that difference is enough to make Dr Bad a murderer, while Dr Good is not guilty of any offence and is regarded as having acted perfectly legitimately, indeed laudably. Why does the difference in the two doctors’ intentions matter so much to the criminal law? We can consider a number of different explanations:

(1) Usurpation. Probably the most historically accurate explanation of why Dr Bad’s intentions make what he did unlawful is that in deciding to kill Patient, he is ‘playing God’ – usurping a decision (as to when Patient will die) that it was for God to make. However, in a secular society, such an explanation is regarded as illegitimate. However, is it possible to argue that Dr Bad’s decision to take Patient’s life was wrongful because it involved Dr Bad in usurping a decision that was for Patient to make? The answer must be ‘no’ because Dr Bad would still be regarded as murderer even if he was simply carrying out Patient’s instructions in injecting him with morphine. We have to look elsewhere for explanations as to why Dr Bad’s intentions make him a wrongdoer.

(2) Patients’ interests. Maybe we could argue that it is in the interests of patients for the law to say that their doctors are not allowed to act with the intention of killing them. The idea is that patients would do worse under any other alternative rule, such as ‘A doctor is allowed to act with the aim of killing one of his or her patients if conditions x, y and z are satisfied.’ Four arguments might be made for saying that patients would be worse under an alternative rule. (i) A slippery slope argument that once we drop the bar on doctors’ intentionally killing their patients by adopting an alternative rule, various institutional and political pressures will result in that alternative rule being gradually reformed and relaxed, and patients will do worse under those successor rules than they would have done had the original bar on intentional killing been retained. (ii) A pessimistic argument that while patients might do better under the alternative rule, if the alternative rule is applied correctly, human error means that the alternative rule will not be applied correctly, and patients will consequently do worse under the alternative rule than they would have done under the original rule which barred intentional killing under any circumstances. (iii) A psychological argument which observes that human beings do not cope well under circumstances of cognitive dissonance – where they are subjected to two conflicting emotions or reactions. In such a situation, human being resolve the dissonance by changing the way they look at the situation that is triggering their emotion or reaction so as to ‘mute’ one of the dissonant emotions or reaction. A doctor who intentionally kills a patient under an alternative rule might well experience a period of cognitive dissonance where his training makes him think ‘It will be a bad thing if this patient dies’ and the law (which gives effect to the alternative rule) makes him think ‘It is okay to bring this patient’s life to an end’. The doctor might well resolve this dissonance by beginning to downgrade the importance of saving his patients’ lives. In other words, he will become desensitised – with sinister implications for his other patients. So the doctor’s patients will get worse treatment from the doctor under the alternative rule than they would have done had the doctor been forbidden from ever intentionally killing a patient. (iv) A combined argument, which says – on the basis of points (ii) and (iii), above – that under an alternative rule, patients will become scared to consult their doctors and will be especially scared to go into hospitals, and will as a result do worse than they would have done had the bar on intentional killing of patients been retained.

I think that all of the arguments (i) – (iv), above, have something going for them, and help to justify the continued existence of a rule barring doctors from intentionally killing their patients, while at the same time doctors are allowed to do things that they know will result in their patients’ dying, so long as doing those things is a necessary and proportionate means of treating their patients’ conditions.

Aggravating intentions

Consider the following two situations. In both cases, Parents have a two-month old Baby.

Useless: Parents neglect to feed Baby because they are spending all their money on drugs; as a result, Baby dies.

Evil: Parents neglect to feed Baby because they resent having to spend any money on her, and hope that she dies; Baby does eventually die.

Parents will be guilty of gross negligence manslaughter in Useless; while they will be guilty of murder in Evil. This is because Parents acted (or, more accurately, failed to act in breach of their duty to feed Baby) with the aim or purpose of killing Baby in Evil, while they had no such aim or purpose in Useless. (I disregard the possibility that Parents might be ‘held’ to have had an intent to kill in Useless under Woollin, on the basis that they foresaw that death was virtually certain to occur. If this troubles anyone, just assume that Parents were so high all the time in Useless that they thought Baby wouldn’t die even if he wasn’t given anything to eat.) Why does the fact that Parents had an intent to kill in Evil make their offence much more serious than the offence committed by the Parents in Useless, who had no such intent? A couple of different explanations can be given:

(1) Character. The Parents in Evil have worse characters than the Parents in Useless. The Parents in Useless did not care enough about the value of their Baby’s life. The Parents in Evil did not care at all about the value of their Baby’s life. This difference in character may lead us to want to stigmatise the Parents in Evil and the Parents in Useless in different ways – by giving them different names (murderers in Evil; manslaughterers in Useless) – so  that people will know what sort of characters they are dealing with when they come across them.

(2) Danger. Because the Parents in Evil have worse characters than the Parents in Useless, the Parents in Evil are more dangerous than the Parents in Useless. The Parents in Evil have proven themselves willing to kill other people, when the Parents in Useless have not provided any evidence that they are willing to do the same. Given this difference in dangerousness, we might want to lock up the Parents in Evil for longer than the Parents in Useless, both to prevent the Parents in Evil doing any more harm to anyone else, and to provide greater opportunities for their rehabilitation.

Why consequences matter to the criminal law

The consequences of your actions matter to the criminal law in a couple of different respects:

(1) The law distinguishes between successful attempts and unsuccessful attempts, so that defendants who are guilty of a successful attempt are treated worse under the criminal law than defendants who are guilty of an unsuccessful attempt – when the difference between a successful and an unsuccessful attempt can just be a matter of luck.

(2) Crimes of constructive liability make a defendant who is already guilty of committing one criminal offence liable for a much more serious offence if his initial crime has certain consequences; and this is so even if the defendant did not intend or foresee that his initial crime would have those consequences. Examples of constructive liability in the criminal law are: (a) murder (where the defendant merely intended to cause gbh to his victim, but death resulted as well); (b) constructive manslaughter; (c) causing death while driving uninsured/disqualified/unlicensed; (d) maliciously inflicting gbh (where all that was foreseen was that the victim would suffer some kind physical harm); (e) assault occasioning actual bodily harm; (f) accomplice liability under the law on joint enterprise.

This post attempts to explain: (1) why the law might distinguish between successful and unsuccessful attempts; and (2) why the law might recognise crimes of constructive liability.

Successful and unsuccessful attempts

Contrast the following two cases:

Hit: Assassin trains his gun on Target, 500 yards away. Assassin pulls the trigger and Target is shot dead.

Miss: Assassin trains his gun on Target, 500 yards away. Assassin pulls the trigger but a sudden gust of wind means that the bullet just misses Target.

In Hit, Assassin will be guilty of murder and get a mandatory life sentence; in Miss, Assassin will only be guilty of attempted murder and won’t get a mandatory life sentence. Why the difference?

A lot of people think that the law should treat Assassin the same in both Hit and Miss as his intentions were exactly the same in both cases, and it’s just a matter of luck that Assassin killed his victim in Hit but not in Miss. I think what this view overlooks is that in both Hit and Miss, Assassin does two things wrong, not one. In both cases, Assassin tries to kill Target – and that’s wrong. But the other thing wrong that Assassin does is that he knowingly exposes Target to the risk of suffering bad luck, by being killed. When Assassin pulls the trigger, he makes it a matter of luck whether or not Target is killed, and he was aware that he was doing that.

So we need to punish Assassin both for (1) trying to kill Target and for (2) knowingly exposing Target to the risk of suffering bad luck. But how do we do this? My suggestion is that we punish Assassin for (1) through a fixed penalty (say, 15 years), and we punish Assassin for (2) through a conditional penalty which Assassin will incur if Target does end up suffering bad luck and is killed. This conditional penalty is a fitting penalty for the wrong committed by Assassin in knowingly exposing Target to the risk of suffering bad luck: if Target ends up suffering bad luck, then Assassin will also suffer bad luck by incurring the conditional penalty.

The result is that we end up punishing Assassin more in Hit than in Miss. In Hit, Assassin incurs both the fixed penalty and also the conditional penalty, as the bad luck that he knowingly exposed Target to the risk of suffering has materialised. But in Miss, Assassin only incurs the fixed penalty: Assassin shares in Target’s good luck – as Target is not killed, Assassin escapes the conditional penalty.

Constructive liability

Contrast the following two cases:

Scary: Man and Wife are driving home from a party, where Man caught Wife kissing another man. Man (who is driving) says to Wife, ‘When we get home, I’m going to beat you to an inch of your life.’ Wife is so scared by this threat that she jumps from the moving car and scars her knees.

Funny: Man and Wife are driving home from a party, where Man caught Wife kissing another man. Man (who is driving) says to Wife, ‘When we get home, I’m going to beat you to an inch of your life.’ Wife laughs and says ‘I’d like to see you try.’

In Scary, Man is guilty of assault occasioning actual bodily harm; while in Funny he is only guilty of an assault. And this is so even though Wife’s reaction to Man’s words may have just been a matter of luck. Why the difference?

We can’t use the ideas discussed in the previous section to explain the difference between Scary and Funny as Man may not have knowingly exposed Wife to the risk of suffering actual bodily harm in either case. (In Scary, he may have been completely astonished that his Wife would react to his threat by jumping from a moving car.) So we need a different idea to explain why the law treats these two cases differently.

We can draw on the idea of a moral threshold to explain why the law might treat Man more harshly in Scary than in Funny.  The idea is that Man crosses a moral (actually, legal – but ‘moral’ sounds better) threshold when he threatens Wife with a beating: he does something wrong. We don’t want him to do this. Now – how do we stop him from crossing that threshold? My suggestion is that by putting Man on notice that if his threat causes his Wife to suffer actual bodily harm, he will be subject to a more serious penalty than he would be if his Wife did not suffer any such harm, the law does two things to help encourage Man to stay the right side of the line of not threatening his Wife:

(1) The law makes Man uncertain as to what punishment he will receive if he crosses the moral threshold and threatens his Wife. He won’t know whether he will get the standard penalty for committing an assault, or a more serious penalty in the event that his assault results in his Wife suffers actual bodily harm. This uncertainty can be expected to deter Man from crossing the moral threshold of threatening his Wife as it makes it much harder for him to determine whether threatening his Wife would be ‘worth it’. (Note that there is no rule of law objection to placing Man in a position of uncertainty as to what punishment he will receive if he assaults his Wife, as Man could easily avoid the uncertainty by simply not assaulting his Wife.)

(2) If Man tries to calculate what the probability is that he will end up receiving a penalty for assault occasioning actual bodily harm, he will be forced to think about what the chances are that his assault will result in his Wife suffering actual bodily harm. In thinking about this, he may be brought to realise what effects his assaulting his Wife will have on her, and will come to the conclusion that given those effects, he should not assault his Wife and will voluntarily decide not to cross the threshold of assaulting his Wife.

So constructive liability may encourage people not to cross the moral thresholds to which constructive liability attaches by: (1) deterring them from crossing those thresholds by making them uncertain what liability they will incur if they cross that threshold; and (2) persuading them not to cross those thresholds by making them reflect on what the effects of crossing those thresholds will be.

The law on intoxication

The basic point

There is no such thing as a defence of intoxication under the criminal law. If you ever find yourself writing ‘D may be able to take advantage of a defence of intoxication’ you are doing something very, very wrong. You need to stop and think – why am I saying this?

Intoxication is only relevant to whether or not a defendant has committed a criminal offence if it means: (1) he did not commit the actus reus of the offence; or (2) he did not have the mens rea for the offence; or (3) he mistakenly believed in facts which, if true, would mean he had a defence to being found guilty of committing that offence. If none of (1), (2) or (3) apply, then the defendant is guilty – the fact that he acted in an intoxicated state will be completely irrelevant.

Kingston illustrates the point. Defendant’s drink was spiked, his inhibitions were consequently loosened, and he ended up sexually assaulting a boy. He knew what he was doing at the time and was in control of his body. He was found guilty: he had the actus reus of the offence (unconsented to touching of a sexual nature), he has the mens rea (intention to touch, or recklessness as to touching) and was not acting under a belief which, if true, would have given him a defence. The fact that the defendant would not have acted as he did had his drink not been spiked was irrelevant. He was still responsible for his actions. If I offered you a million pounds to kill your best friend, and you took the money and killed your best friend, you could not dream of trying to get off the charge by arguing, ‘Had I not been offered the money, I would never have killed my best friend.’ And yet people still try to argue that Kingston should have been acquitted. Whatever it was that caused you to loosen your inhibitions – money, drugs, drink, lust – if you are still responsible for your actions, you can still be held accountable for them under the criminal law.

Let’s now look at how the law handles cases when a defendant can argue that either (1), (2) or (3), above, apply.

No actus reus

In this sort of case, the effect of D’s intoxication either means that D had no idea what he was doing, or D was not in control of his body. In either case, D might be able to argue: I didn’t have the actus reus because I didn’t cause whatever harm it is that my body did while I was blacked out or had no control over my body’s physical movements. If this argument works, then D is entitled to be acquitted on the basis of non-insane automatism (not insane automatism, because the automatism had an external cause – the intoxication). But there is one exception. If D was responsible for his own intoxication then we can say (at least where the actus reus of the offence requires that it be shown that D caused a particular harm) that D did have the actus reus. By causing himself to get drunk, he was the ultimate cause of the harm that his body did while D was blacked out or not in control of his body’s physical movements. So we can contrast two cases:

Wild Thing 1. D gets so drunk in a pub that he ends up having convulsions. As he is flailing around, one of his fists hits a fellow drinker, V. In this case D has the actus reus of assault occasioning actual bodily harm (assuming that the hit did abh to V). D  ultimately caused his fist to touch V by getting so drunk that he went into convulsions, and this caused (we are supposing) V to suffer actual bodily harm.

Wild Thing 2. Same as Wild Thing 1 except D was drinking what he thought was orange juice, but it had been spiked by T. In this case, D does not have the actus reus for assault occasioning actual bodily harm because he did not cause his fist to hit V. T did that; so if anyone is going to be guilty of assault occasioning actual bodily harm, it should be T. But D is entitled to be acquitted here on the basis of non-insane automatism.

No mens rea

In this sort of case, the effect of D’s intoxication means that D can deny that he had the mens rea for the offence with which he has been charged. For example, if, say, he was charged with murder, D would be arguing – my intoxicated state means that I did not have an intent to kill or an intent to cause gbh. If, on the other hand, he was charged with malicious wounding, D would be arguing – I did not appreciate that there was an unreaonable risk that V would suffer some physical harm from my actions. Here are a couple of situations where D would be able to make such an argument:

Polar Bear. D gets blind drunk in a pub. As he walks outside the pub, alone, he sees what he thinks is a polar bear coming towards him. (In fact, it is V, dressed in a white parka.) Thinking that the polar bear’s head would look great on his wall, D grabs hold of what he thinks is the polar bear and cuts its head off with a handy knife. On sobering up, D is told that he decapitated V and is being charged with murder. D has the actus reus for murder: he caused V’s death by sawing V’s head off. But D will argue that he did not have the mens rea for murder when he was sawing V’s head off: he intended to kill a polar bear, not a human life in being.

Note that this argument would not work if D was getting blind drunk in order to get up the courage to go home and kill his wife. In such a case, we could dial back the moment when D caused V’s death to the moment when he was getting drunk in the pub, and say that that act of getting drunk caused V’s death because V would not have died had D not gotten blind drunk, and there was no obvious break in the chain of causation after the moment he got drunk to break the chain between the getting drunk and V’s dying. And at the moment D got drunk, he did have an intent to kill – he was getting drunk with the aim or purpose of killing his wife. So in that situation, D would not be able to deny that he had the mens rea for murder, contemporaneous with the actus reus, and would consequently be convicted of murder. This is known as the ‘Dutch courage rule’ (first articulated by Lord Denning in Attorney-General for Northern Ireland v Gallagher).

But let’s assume the ‘Dutch courage rule’ doesn’t apply in Polar Bear: D was getting drunk just for the hell of it. In such a case, D would argue that he should be acquitted because he did not have the mens rea for murder.

Darts. D gets blind drunk in a pub. He gets up, intending to have a game of darts. But he is so drunk, he thinks that the darts in his hand are actually little wriggling mice. He is so disgusted, he throws one of the ‘mice’ at a nearby drinker, V, shouting, ‘Look at this! That’s horrible!’ When D sobers up, he is told that V was actually struck in the eye by the dart that he threw at V, and that D is going to be charged with malicious wounding, contrary to s 20 of the Offences Against the Person Act 1861. D has the actus reus: what he did in throwing the ‘mouse’ at V caused V to be wounded. But D will argue that he did not have the mens rea: when he threw the ‘mouse’ at V, he did not foresee that V would suffer ‘some kind of physical harm’ as a result of his actions.

In cases where D seeks to rely on his own intoxicated state in order to deny that he had the mens rea of the offence with which he has been charged, a very simple rule (called the Majewski rule) applies to determine whether D can do this. In the case where D was to blame for getting drunk, the courts ask – Had D not been drunk, would he have had the mens rea for the offence with which he has been charged? If the answer is ‘yes’ then D will not be allowed to deny that he had the mens rea for the offence with which he has been charged. If the answer is ‘no’, then D will be allowed to persist in his plea that he did not have the mens rea for the offence with which he has been charged, and will be acquitted on the basis of no mens rea.

Let’s see how the Majewski rule would apply in our two situations:

Polar Bear. Here D’s drunken state means that he did not have the mens rea for murder, but he was also to blame for getting drunk, so we might prevent him denying that he had the mens rea for murder depending on the application of the Majewski rule. So we ask – Had D not been drunk, would he have had an intent to kill or an intent to cause gbh? The answer, pretty obviously, is ‘no’ so D is allowed to persist in his plea that he did not have the mens rea for murder and will be acquitted on the basis of no mens rea.

Darts. Here D’s drunken state means that he did not have the mens rea for malicious wounding, but he was also to blame for getting drunk, so we might prevent him denying that he had the mens rea for malicious wounding depending on the application of the Majewski rule. So we ask – Had D not been drunk, would he have foreseen that there was a risk of V suffering some kind of physical harm when he acted as he did? Well, had D not been drunk, he would have realised he was holding a dart, not a mouse, so he plainly would have foreseen that there was a risk of V suffering some kind of physical harm when he threw the dart at V. So the answer to our question is ‘yes’ and as a result D will not be allowed to deny that he had the mens rea for malicious wounding. As a result, D will be convicted of malicious wounding – he committed the actus reus, and is not allowed to deny that he had the mens rea, and will have no relevant defence.

The textbooks try to sum up the above by saying that a defendant who is charged with a criminal offence will – if he was to blame for getting drunk – be allowed to rely on his drunken state to deny that he had the mens rea for the offence if the offence was one of ‘specific intent’ (= mens rea requires an intent to produce a particular outcome), but not if the offence was one of ‘basic intent’ (= mens rea can be established merely by showing that the defendant foresaw that his actions would have a particular outcome). But this is dangerously vague. It would be better just to ask, in a case where the defendant seeks to rely on his drunken state to establish that he did not have the mens rea for the offence with which he has been charged, and the defendant was to blame for getting drunk, the Majewski question: Would the defendant have had the mens rea had he not been drunk? Answering that question will allow you to determine whether or not the defendant will be allowed to deny that he had the mens rea without having to worry about whether the offence that the defendant has been charged with is one of ‘specific intent’ or ‘basic intent’.

Mistaken belief in defence

In this sort of case, D’s intoxicated state means that he believed facts were true, that if they were true would have given him a defence, such as self-defence, or duress. Examples:

Banana Man. D is high on LSD. In his drunken state, he believes V is running towards him with a knife, and is about to stab him. D takes out a gun and shoots V dead. In fact, V was not holding a knife but a banana, which he was urging D to eat.

Down at Heel. D is in the pub with V. V is telling D about his financial problems, and says to D, ‘You’ll take care of me, won’t you? You’d really regret it if you didn’t.’ D is pretty drunk and misinterprets V as threatening that he will kill D unless D immediately gives him money. D pulls out a gun and holds up the pub, netting £2,000, which he instantly hands over to an astonished V.

In these cases, the courts ask a very simple question – Was D to blame for his intoxicated state? If he was, then he will not be allowed to rely on his mistaken belief to argue that he has a defence. But if he was not to blame for being intoxicated, and if the facts that D’s intoxicated state led him to believe were true would – if true – have given him a defence, then D will be allowed to raise that defence to the offence with which he has been charged.

So if, in Banana Man, someone had slipped D a tab of LSD without his realising it, and shooting V would have been a reasonable thing to do had V actually been about to knife D, then D will be acquitted of murder on the ground of self-defence. (Note that in Banana Man, the fact that D had the actus reus and mens rea for murder is easily established.) But if in Banana Man, D had taken the LSD, he will not be allowed to rely on his mistaken belief to argue that he should be acquitted on the ground of self-defence. (Note that this is an exception to the normal approach, established in Gladstone Williams, that a defendant will be entitled to raise a defence of self-defence so long as he honestly believed in facts that, if true, would have given him a defence of self-defence.)

The same is true in Down at Heel. If D was to blame for getting drunk, he will not be allowed to rely on a defence of duress when he is charged with theft. But if someone had been spiking his orange juice, he will be allowed to rely on his mistaken belief that V was threatening to kill him to set up a defence of duress. (This is assuming that on the facts that he believed to be true he would have had a defence of duress.)

The one case where a defendant who was to blame for getting drunk will still be allowed to rely on a mistaken belief in facts which, if true, would have given her a defence was established in Jaggard v Dickinson. Under s 5(2) of the Criminal Damage Act 1971, it is a defence to being convicted of causing criminal damage to property belonging to another if you believed the owner of the property would have consented to the damage. The defendant, in a drunken state that she was to blame for getting into, attempted to break into her friend’s house believing that her friend would be happy for her to do this. In fact, she had got the wrong house and was breaking down someone else’s door. She was acquitted of criminal damage to property on the basis that she honestly believed that the owner of the door would have been happy for her to break down the door. This was so even though the only reason she thought this was because she was drunk, and she was to blame for getting drunk.

Criticisms of the law

The law on intoxication is sometimes criticised on the basis that –

(1) It is unclear.

(2) It does not provide a defence to someone like the defendant in Kingston.

I don’t think either of these criticisms are justified. The law is pretty clear (as you can see from my summary above), and there is no way the law should be acquitting the defendant in Kingston. But this does not mean that the law in this area is beyond criticism. Two criticisms can be made:

(3) In Darts we end up convicting the defendant of a very serious offence – malicious wounding – simply on the basis that he was to blame for getting drunk. This does not seem right. If the defendant, when he got drunk, did not foresee the risk of other people coming to some physical harm as a result, he is being found guilty of an offence for which such foresight is required when he had no such foresight. He is being found guilty because he is prevented, under the Majewski rule, from denying that he had such foresight. Convicting someone on the basis of a fiction seems very unsatisfactory. It looks like the defendant in Darts is being convicted on a constructive basis – he did something bad when he got drunk, and then the offence he ends up being convicted of depends on the level of harm that resulted from his getting drunk without any requirement that he needs to have foreseen that harm.

(4) The workings of the above rules seems to produce some arbitrary results. Consider:

Tramp 1. Two tramps, Vladimir and Estragon, get very drunk and take a lot of drugs and fall asleep near each other. Vladimir wakes up, still intoxicated, and thinks that the sleeping Estragon beside him is a writhing mass of snakes that are about to attack him. Vladimir panics, pours the remaining alcohol near him onto Estragon and sets it on fire. Estragon dies. Vladimir is charged with murder but is acquitted on the basis that he lacked the mens rea for murder: he did not have an intent to kill a life in being. See R v Lipman.

Tramp 2. The same facts as Tramp 1, but in this case when Vladimir wakes up, he wrongly thinks that Estragon is attacking him with a knife. Vladimir strangles Estragon. Vladimir is charged with murder and is convicted on the basis that he had the actus reus for murder (he caused Estragon’s death), and the mens rea for murder (he intended to kill Estragon), and he can’t rely on his mistaken belief that he thought that Estragon was attacking him because that mistaken belief was the result of his intoxicated state, and he was to blame for being intoxicated. See R v O’Grady; R v Hatton.

But what, morally, is the difference between Tramp 1 and Tramp 2 that would justify us in acquitting Tramp 1 of murder, and finding Tramp 2 guilty? The treatment of Tramp 2 looks even more arbitrary given the fact that under Gladstone Williams a non-intoxicated but still stupid belief that one is about to be attacked might entitle a defendant to be acquitted of murder.

The defence of necessity

What is it?

A defendant seeks to rely on a necessity-type defence when he argues that he should not be convicted of a criminal offence because his conduct was justified, and it was justified because he did more good than harm in acting as he did.

Defences that are not necessity defences

Given this definition, two defences that are often categorised as being instances of an overarching defence of ‘necessity’ actually having nothing to do with necessity, properly understood.

(1) Duress

The defence of duress – which is available to a defendant who acted under the pressure created by the need to avoid death or serious injury to himself or to someone for whom he is responsible – cannot be analysed as a necessity defence because a defendant seeking to raise a defence of duress is not seeking to justify his conduct. Instead he is seeking to excuse his conduct by arguing, in essence, that ‘I know what I did was wrong, but I acted as I did under pressure that would have led a normal person to act in the same way as I did. So it’s unfair to find me guilty of committing an offence here because had you been in the same position as me, you would have done the same as me. So it’s just an arbitrary accident of fate that had led me rather than you to end up being charged with this offence, and we shouldn’t allow arbitrary accidents of fate to determine people’s criminal guilt.’

(2) Self-defence

The defence of self-defence is often analysed as being a form of necessity defence because: (1) a defendant who seeks to rely on the defence of self-defence is seeking to show that his conduct was justified, just as he would be doing if he were seeking to rely on a defence of necessity; and (2) a defendant who seeks to rely on the defence of self-defence will not be allowed to rely on the defence unless he used reasonable force to defend himself given the facts as he believed them to be, just as a defendant who was seeking to rely on a necessity defence would not be allowed to unless he could show that his actions were reasonable, all things considered.

However, the better analysis seems to be that in a self-defence case, the defendant is arguing that his conduct was justified not because he did more good than harm in using force against his attacker, but because his attacker had no right that the defendant not use force against him. For example, suppose that Anna is being raped by David when Anna manages to reach for a knife and stabs David to death. Most people agree that Anna is entitled to be acquitted of murder here: she will be able to take advantage of a defence of self-defence. But that is not because she did more good than harm in stabbing David to death. Whether or not that was the case is irrelevant. Anna’s conduct here was justified because David had no right that Anna not stab him to death, given what he was doing to Anna. By raping Anna, David forfeited the right he would have normally had against her that she not subject him to lethal force.

(Note that this analysis does not work to explain cases like R v Gladstone Williams (1984) where V was not actually attacking D, but D used force against V because she honestly believed V was a threat to her; in such a case, it is hard to say that V had no right that D not use force against him. Gladstone Williams must be explained on some other basis, such as that D’s conduct can be excused. But a problem with this is that in an excuse case, the defendant has to show that his conduct was reasonable, so as to allow the defendant to argue that ‘You would have done the same as me had you been in my position’ – and the effect of Gladstone Williams is to acquit a defendant who has acted on the unreasonable belief that he or she is being attacked.)

Statutory defences that look like necessity defences

There are some statutory defences that, it could be argued, look like forms of necessity defences:

(1) Damage to property

Under s 5(2) of the Criminal Damage Act 1971, a defendant who is charged with causing criminal damage to another’s property will have a defence if he acted as he did in order to protect other property from being damaged, and he acted reasonably in that belief. The classic example where this defence would apply where a fire is raging down a row of houses; the fire services will be authorised to pull down a house in the middle of the row so as to create a fire break and stop the fire spreading any further. By pulling down the house in the middle of the row, the fire fighters have done more good than harm and are entitled to justify their conduct on that basis.

(2) Refugees

Under s 31 of the Immigration and Asylum Act 1999, a defendant who has been charged with committing various offences in order to enter the UK will have a defence if he came to the UK directly from a country ‘where his life or freedom was threatened’ and promptly presented himself to the authorities and made a claim for asylum. A defendant who prevented himself being killed or imprisoned by entering the UK under a false passport, or by smuggling himself across the UK’s borders, will have done more good than harm and is entitled to justify his conduct on that basis.

(3) Crime stoppers

Section 3(1) of the Criminal Law Act 1967 says that ‘A person may use such force as is reasonable in the circumstances in the prevention of crime, or in effecting or assisting in the lawful arrest of offenders or suspected offenders or of persons unlawfully at large.’ Provisions of the Police and Criminal Evidence Act 1984 confer numerous powers on the police to do things that would otherwise be criminal offences, such as stopping and searching someone (s 1), seizing property (s 19), arresting someone (s 24) and detaining someone in custody (Part IV). Such provisions can most easily be explained on the basis that the sort of conduct covered by these provisions is in the public interest and is therefore justified.

Necessity under the common law

It seems that the courts do not accept that a defendant who has been charged with a criminal offence can ever take advantage of a defence of necessity under the common law. So unless some statute provides the defendant with a necessity-type defence, it will not do the defendant any good to plead, ‘In acting as I did, I did more good than harm.’ Such a plea will fall on deaf ears.

This may seem a controversial claim, given the many cases in which the courts have begun to recognise the existence of a general defence of necessity and set out the situations in which such a defence will be available to a defendant. However, in all those cases (except two) the courts consistently run together the defence of necessity with the defence of duress of circumstances. They do this so consistently that when the courts talk of a defendant being allowed to take advantage of a defence of ‘necessity’ it seems that they are really talking about when a defendant will be allowed to take advantage of a defence of duress of circumstances.

Consider, for example, the following quote from Simon Brown J in R v Martin (1989) (driving while disqualified):

‘First, English law does, in extreme circumstances, recognise a defence of necessity. Most commonly this defence arises as duress, that is pressure upon the accused’s will from the wrongful threats or violence of another. Equally, however, it can arise from other objective dangers threatening the accused or others. Arising thus it is conveniently called “duress of circumstances”.

‘Secondly, the defence is available only if, from an objective standpoint, the accused can be said to be acting reasonably and proportionately in order to avoid a threat of death or serious injury.

‘Thirdly, assuming the defence to be open to the accused on his account of the facts, the issue should be left to the jury, who should be directed to determine these two questions: first, was the accused, or may he have been, impelled to act as he did because as a result of what he reasonably believed to be the situation he had good cause to fear that otherwise death or serious physical injury would result? Second, if so, may a sober person of reasonable firmness, sharing the characteristics of the accused, have responded to that situation by acting as the accused acted? If the answer to both those questions was yes, then the jury would acquit: the defence of necessity would have been established.’

This passage was endorsed in R v Abdul-Hussain (1998) (hijacking of plane) as providing ‘the clearest and most authoritative guide to the relevant principles’ as to when a defence of necessity would be available. (Which dictum was in turn approved in R v Shayler (2001) (breach of Official Secrets Act 1989) and R v S (2012) (child abduction).) But it is clear that Simon Brown J was not talking about necessity at all – in the sense in which I have defined that defence – but was rather talking about when a defendant will be able to take advantage of a defence of duress (either duress by threats or duress of circumstances). As we have already observed, duress is not necessity. As Brooke LJ observed in In Re A (2001) (separation of conjoined twins), ‘In cases of pure necessity the actor’s mind is not [as it is in duress cases] irresistibly overborne by external pressures. The claim is that his or her conduct was not harmful because on a choice of two evils the choice of avoiding the greater harm was justified.’

Despite this, the courts have continued to identify the defence of necessity with the defence of duress of circumstances. In R v Conway (1989) (reckless driving), Woolf LJ held that:

‘necessity can only be a defence to a charge of reckless driving where the facts establish “duress of circumstances,”… i.e. where the defendant was constrained by circumstances to drive as he did to avoid death or serious bodily harm to himself or some other person.  …Whether “duress of circumstances” is called “duress” or “necessity” does not matter. What is important is that, whatever it is called, it is subject to the same limitations as [the] “do this or else” species of duress.’

Twelve years on, in R v Shayler (2001, CA), Lord Woolf CJ was still of the same mind. Dismissing arguments that necessity and duress were distinct in that necessity is a justification and duress is an excuse, he said: ‘…the distinction between duress of circumstances and necessity has, correctly, been by and large ignored or blurred by the courts… the law has tended to treat duress of circumstances and necessity as one and the same’ (at [55]).

The two cases that might genuinely be said to have recognised the existence of a defence at necessity under the common law are Gillick v West Norfolk and Wisbech AHA (1986) and In Re A (2001). Let’s look at each of these in turn.

In Gillick, the issue was whether a doctor who prescribed contraception to a girl who under 16 would be guilty of an offence under s 28 of the Sexual Offences Act 1956 of causing or encouraging unlawful sexual intercourse ‘of a girl under the age of sixteen for whom he is responsible’. (Note that this offence was repealed by the Sexual Offences Act 2003, s 10 of which makes it an offence to cause or incite someone who is under 16 to engage in a sexual activity.) The majority (Lords Fraser, Scarman and Bridge) took the view that a doctor may well commit an offence under s 28 if he prescribed contraception to an under 16 year old girl ‘with the intention of facilitating her having unlawful sexual intercourse’, but a doctor could not be said to have such an intention if he was honestly prescribing the contraception ‘for the maintenance or restoration of [the girl’s] health.’ (Both quotes from Lord Scarman’s judgment.) The minority (Lords Templeman and Brandon) disagreed, holding that if a doctor prescribed contraception to a girl who was under 16, he would be ‘promoting, encouraging or facilitating the having of sexual intercourse’ (per Lord Brandon) and as a result would commit an offence under s 28. None of the judgments in Gillick say anything about necessity: they were purely concerned with the question of how s 28 of the 1956 Act should be interpreted. So there is nothing in Gillick that can be said to support the existence of a defence of necessity under the common law.

That leaves In Re A, in which the Court of Appeal had to decide whether it would be lawful to carry out an operation to separate two conjoined twins, Jodie and Mary, even though such an operation would inevitably result in the death of one of the twins, Mary. There was no possibility of arguing that the defence of duress of circumstances could apply to make the operation lawful: (1) duress cannot be excuse murder (a defendant cannot claim ‘you would have done the same as me’ when what he has done is intentionally kill someone else); (2) as Robert Walker LJ observed ‘The doctors are not faced with any threat to themselves, but they are faced with the anxious dilemma of trying to perform their professional duties which they owe to their two infant patients.’ So In Re A provided an excellent test case for whether the Court of Appeal were willing to recognise a defence of necessity independent of a defence of duress of circumstances.

All three judges in the Court of Appeal held that the doctors would not be committing a crime in operating to separate the conjoined twins.

Ward LJ held that the operation would be justified on the grounds of self-defence, as Mary – who was dependent on being supplied with blood from Jodie’s heart to stay alive, with the result that Jodie’s heart was destined to fail if she remained joined to Mary – posed an ‘unjust’ threat to Jodie’s life.

Brooke LJ held that the operation could be justified on the ground of necessity. He held that ‘there are three necessary requirements for the application of the doctrine of necessity: (i) the act is needed to avoid inevitable and irreparable evil; (ii) no more should be done than is reasonably necessary for the purpose to be achieved; (iii) the evil inflicted must not be disproportionate to the evil avoided.’ He thought that all three conditions were satisfied in this case. It is not, however, clear that Brooke LJ thought that satisfying conditions (i) – (iii) would be sufficient to give rise to a defence of necessity. A few paragraphs before setting out these conditions, he held that there were two objections to recognising a defence of necessity in a case where a defendant was charged with committing a criminal offence: (1) that it was beyond the competence of the courts to assess whether the operation would do more good than harm; (2) that allowing a defence of necessity would result in the court authorising someone to act immorally. Brooke LJ thought that neither objection applied In Re A. On (1), it was obvious whether or not the operation would do more good than harm: dealing with this question did not require the courts to put a comparative value on Jodie and Mary’s lives. If the operation were not carried out, both Jodie and Mary would die. If it were carried out, Jodie would live and Mary would die. Given this, it was obvious what was the better option. On (2), it was not clear that carrying out the operation would be an immoral act, so it was not clear that by allowing the operation to be carried out the court would be ‘marking an absolute divorce of law from morality’.

Robert Walker LJ seemed to take the view that the operation did not even approach the threshold of criminality that would require the doctors to rely on a defence to justify or excuse their actions. He held that the doctors carrying out the operation could not be said to have intended to kill Mary (so they would not have the mens rea for murder), and could be said to be acting in Mary’s best interests in carrying out the operation as ‘The operation would give her, even in death, bodily integrity as a human being’ (so they would not be breaching the duty of care they owed Mary as her doctors in carrying out the operation and could not therefore be held guilty of gross negligence manslaughter).

So – of the three judges who decided In Re A, only one (Brooke LJ) rested his decision on the basis that the doctors could take advantage of a defence of necessity which was genuinely distinct from a defence of duress of circumstances. And Brooke LJ’s judgment remains the only English authority that there exists a defence of necessity at common law that is distinct from a defence of duress of circumstances. Given this, it seems fair enough to say that the English law does not currently recognise that a defendant who has been charged with a criminal offence can ever rely on a defence of necessity to justify his conduct in the absence of some specific statutory provision which allows him to raise a necessity-type defence to the charge.

Why are the courts so unwilling to recognise a free-standing defence of necessity?

There seem to be four reasons for the lack of authority in favour of the proposition that a defendant is entitled to be acquitted of a criminal charge if he can show that he did more good than harm in acting as he did.

(1) The definition of offences

There are two ways in which the definition of an offence can exclude a defendant from seeking to rely on a defence of necessity to justify his conduct.

First, the definition of an offence may make it unnecessary for the defendant to rely on a defence of necessity to justify his conduct, because the facts which the defendant might have wanted to rely on as giving him a defence of necessity will also allow him to argue that he does not commit the actus reus of the offence with which he is charged, or that he did not commit the actus reus of the offence with the requisite mens rea. For example, in R v Abdul-Hussain (1998), Rose LJ observed that ‘if Anne Frank had stolen a car to escape from Amsterdam and been charged with theft, the tenets of English law would not have denied her a defence of duress of circumstances, on the ground that she should have waited for the Gestapo’s knock on the door [before attempting to escape in a stolen car].’ But in such a case Anne Frank would not have needed to rely on a defence to avoid being convicted of theft. She would have been able to argue that in the circumstances, her appropriation of the car was not dishonest – with the result that she did not commit the actus reus or mens rea of theft (depending on how you classify the requirement of dishonesty in theft, as going to the actus reus or the mens rea of theft).

Secondly, where an offence has been created under a statute, it might be impossible for the defendant to rely on a defence of necessity to justify his conduct because it will be clear from the definition of the offence that Parliament did not intend such a defendant to be allowed to escape conviction by pleading that his actions were in the public interest. The most obvious example of this is the Official Secrets Act 1989. The Act does not provide any defence to someone who claims that they were acting in the public interest in making public secrets ‘relating to security or intelligence’ that they came across by virtue of their position as ‘a member of the security and intelligence services’ or as a government official who has been notified that they are bound by the Act. Given this, the courts would be flouting the will of Parliament if they were to allow such a defendant a defence of necessity under the common law; and they so ruled in the case of R v Shayler (2003, HL). For similar reasons, the appeal in R v Quayle (2005) – where the defendant attempted to justify their possession of marijuana on the basis of ‘medical necessity’ – was doomed to fail. The Misuse of Drugs Act 1971 provides no defence to those who possess marijuana for medical purposes. Given the extreme likelihood that this was a deliberate decision on the part of Parliament, it would have been a violation of Parliamentary sovereignty for the courts to hold that the defendants in Quayle were entitled to a defence of necessity. It was different in DPP v Pipe (2012), where the defendant was charged with breaking the speed limit when he drove his son – who had broken his leg playing football – to hospital. Allowing him to rely on a defence of necessity (here, a synonym for ‘duress of circumstances’) did not involve the courts in violating the will of Parliament in setting speed limits for motorists. Parliament – or the ministers exercising powers to set speed limits delegated to them by Parliament – could not be expected to have had circumstances such as those in DPP v Pipe in mind in setting speed limits.

(2) Rights

Most common law offences exist to protect individual’s rights that other people not treat them in certain ways (kill them, rape them, take their property, batter them, and so on). The fact that English law recognises that people have rights reflects the individualistic ethos of English law, under which greater priority is placed on protecting the freedom of the individual than the welfare of the community. Rights give people a veto power over what can be done to them in the name of the public interest. In Ronald Dworkin’s vivid phrase, ‘rights are trumps’.

In contrast, the existence of necessity defences reflect a more communitarian ethos, under which greater priority is placed on promoting the welfare or the community than on protecting the freedom of the individual. If ‘I did more good than harm’ were always a defence to being charged with committing a criminal offence, then people would have no rights. All people would have were interests that would be taken into account in determining what the right thing to do would be. So if V’s interests in not being killed by D were outweighed by the community’s interests in seeing that V was killed by D, then the law would permit V to be killed by D. It would not be possible anymore to say that V had a right not to be killed by D because killing V might be allowed, depending on the circumstances.

Of course, if the balance of interests indicated that V should not be killed, then we could say in that situation that V has a right not to be killed by D. But that statement, that V has a right not to be killed by D, would simply represent the conclusion of a chain of reasoning: ‘It would do more harm than good to kill V, so V should not be killed, so V has a right not to be killed by D.’ The statement that V has a right not to be killed by D would not – as it does in a community which adopts an individualistic ethos – form the premise of a chain of reasoning: ‘V has a right not to be killed by D, so it would be wrong for D to kill V even though killing V would, on balance, do more good than harm.’

It follows that so long as English law adopts an individualistic ethos, it cannot recognise the existence of necessity defences to crimes which involve violating an individual’s rights. Recognising the existence of such defences would be fatal to English law’s individualistic ethos, in the same way that kryptonite is fatal to Supeman: they cannot co-exist in the same space. English law could not at one and the same time recognise that people have rights and recognise that a defendant could justify treating someone in a particular way by arguing ‘I did more good than harm.’ It has to be one or the other; not both. Once we grasp this, we can see that arguments about whether the law should recognise a defence of necessity (in the sense that it has been defined in this essay, a justification based on the plea ‘I did more good than harm’) are really debates about whether the law should be influenced by an individualistic ethos or by a communitarian ethos. And the fashionability of communitarian ideas among academics explains why so many academics are in favour of the criminal law recognising a fully-fledged defence of necessity.

(3) Judgment

Even if English law were to adopt a more communitarian ethos – and many people would say that it has been in the process of doing that since the 1970s – the courts would still have good reason to hesitate over whether or not to allow defendants to rely on a necessity defence, at least in cases where they are not excluded from doing so by the terms of the offence with which they have been charged. In a case where a defendant argued ‘My conduct was justified because I did more good than harm’, it will often be very difficult for the courts to judge whether or not the defendant’s argument is made out. The problem is not really that the courts lack the expertise to weigh the goods and harms that were produced by the defendant’s conduct, but more that no one could weigh those goods and harms because no one can tell what those goods and harms are. For example, in R v Dudley and Stephens (1884), the court held that the defendant seaman – who had abandoned ship in the middle of a storm and put to sea in an open boat with a cabin boy – could not justify killing the cabin boy in order to eat his body on the basis that had they not done so, they would have all starved to death. Lord Coleridge CJ pointed out that even if it were the case (which was not actually certain, as they might have been picked up any day by a passing ship; or they might not have been picked up at all, ever) that the people in the open boat had a choice between ‘One of us dies, and the rest of us live’ and ‘We all die’, that could not justify killing the cabin boy as opposed to any of the other people in the boat. Killing the cabin boy could only be justified if that was the least worst option. But:

‘Who is to be the judge of this sort of necessity? By what measure is the comparative value of lives to be measured ? Is it to be strength, or intellect, or what? …In this case the weakest, the youngest, the most unresisting, was chosen. Was it more necessary to kill him than one of the grown men? The answer must be “No”…’

None of us can really tell what the consequences of our actions will be. Given this, asking the courts to determine whether the defendant’s actions did more good than harm is asking them to do the impossible. Even in a case like In Re A, where it seems clear-cut that going ahead with the operation would do more good than harm, it is not impossible to imagine futures where not carrying out the operation would have been the better option: for example, a future where Jodie grew up to become a mass murderer, or a future where Jodie became so depressed at owing her life to the death of her twin that she was never able to function normally or enjoy life. And it is not impossible to imagine twists on those futures that would allow us to say that, in fact, carrying out the operation was actually still the better option, where the tragedies resulting from Jodie’s existence had outcomes that worked to offset the initial suffering involved in those tragedies.

(4) Disorder

The final reason for being cautious about recognising the existence of necessity defences, even in a legal system that is fully committed to adopting a communitarian ethos, is the potential for disorder that the existence of such defences creates. The criminal law will speak with a more uncertain voice, and provide less of a constraint on people’s actions, if people know that their conduct will not be attract a criminal sanction if they can establish that they did more good than harm in acting as they did. This consideration played some part in the decision of the court in R v Dudley and Stephens to deny a necessity defence to the defendants in that case. Lord Coleridge CJ warned of the ‘awful danger’ involved in allowing that ‘the temptation to murder’ could give rise to a defence to a charge of murder:

‘such a principle once admitted might be made the legal cloak for unbridled passion and atrocious crime. There is no safe path for judges to tread but to ascertain the law to the best of their ability and to declare it according to their judgment… [A judge] has no right to declare temptation to be an excuse, though he might himself have yielded to it, nor allow compassion for the criminal to change or weaken in any manner the legal definition of the crime.’

Almost a hundred years later, in Southwark London Borough Council v Williams (1971), Lord Denning MR advanced similar reasons for refusing to find that squatting in abandoned premises could be justified on the grounds of necessity:

‘The doctrine [that private property may be sacrificed to avoid ‘great and imminent danger’] must…be carefully circumscribed. Else necessity would open the door to many an excuse. It was for this reason that it was not admitted in Reg. v. Dudley and Stephens (1884) 14 Q.B.D. 273, where the three shipwrecked sailors, in extreme despair, killed the cabin boy and ate him to save their own lives. They were held guilty of murder. The killing was not justified by necessity. Similarly, when a man, who is starving, enters a house and takes food in order to keep himself alive. Our English law does not admit the defence of necessity. It holds him guilty of larceny. Lord Hale said that “if a person, being under necessity for want of victuals, or clothes, shall upon that account clandestinely, and animo furandi, steal another man’s food, it is felony…”: Hale, Pleas of Crown, i. 54. The reason is because, if hunger were once allowed to be an excuse for stealing, it would open a way through which all kinds of disorder and lawlessness would pass. So here. If homelessness were once admitted as a defence to trespass, no one’s house could be safe. Necessity would open a door which no man could shut. It would not only be those in extreme need who would enter. There would be others who would imagine that they were in need, or would invent a need, so as to gain entry. Each man would say his need was greater than the next man’s. The plea would be an excuse for all sorts of wrongdoing. So the courts must, for the sake of law and order, take a firm stand. They must refuse to admit the plea of necessity to the hungry and the homeless: and trust that their distress will be relieved by the charitable and the good.’

Conclusion

It seems that the current state of law on when a defence of necessity (properly understood as a justification resting on the argument ‘I did more good than harm’) will be available is justified. If necessity-type defences are to be recognised under the law, it would be better if they were created by Parliament, rather than the courts. Giving effect to statutory necessity defences would pose no threat to the doctrine of Parliamentary sovereignty. Moreover, Parliament is in a better position than the courts to judge when acting in a particular way can be justified on the basis that acting in that way did more good than harm, and to set out what factors are to be taken into account in reaching that conclusion. Further, the statutory necessity-type defences are less likely to be abused or misused as their scope should be clearly delineated and understood in advance; something that is not possible when the courts undertake to sketch out on a case-by-case basis when they will permit a defendant to rely on a necessity defence.