Can the Chain of Causation ever be broken in cases of Manslaughter?

The simple maxim that the accused in a criminal proceeding ought to have caused the crime is one of the potentially most important doctrines in law. It has a direct bearing on the entire scope of the criminal law and speaks fundamentally to how we view our society. The proximity between cause and effect that we consider to bring rise to a positive moral duty of care, when we refer to the imposition of punitive sanctions on members of society, is one of the most important tests of criminality.

Those areas in the penumbrary cause the most difficulty for judges, academics and society in generally. Therefore the prosecution of drug dealers for the death of those they supply , doctors who were negligent and contribute to the death of their patients , Reckless transmission of HIV / Aids to sexual partners  or the denial of access to life-saving treatment  are all areas where the person being accused of the crime may or may not be considered to have ‘caused’ the harm. The quotation above by Seneca represents one of the key difficulties in establishing a chain of causation from the accused to the victim of any crime. In English law omission to act cannot be considered the cause of a crime because there is not the causal link; however there is an undoubted argument that someone who refuses to help upon seeing a crime is encouraging that action.

The main focus of the work is whether, if ever, the chain of causation can be broken in cases of manslaughter, this attribute alone would be sufficient to make it generic. The work starts with a section on the general principle of causation in criminal law. We will then move onto have an in depth discussion and analysis concerning the case law on the issue before finally using the other jurisdictions and areas of law to compare to the application of the test in manslaughter. This is an incredibly vast subject matter and the restriction to consideration of manslaughter is hoped to give the essay more poignancy and give a more holistic understanding to the operation of causation from which we maybe able to extrapolate general principles about causation. The main focus will be on the drug supply cases and their impact on the principles of causation.  

1: The Concept of Causation

Causation is an extremely important part of the Actus Reus in result crimes. The failure of the prosecutor to establish causation is disastrous to a case because ‘if the death, damage or destruction occurred because of some other cause then the offence has not been committed even though all the other elements of the actus reus are present and the accused had the necessary mens rea’   However, it is far from a simplistic concept of deciding that the action caused the result. The concept of causation represents a nexus between fact and law, they are at certain points clearly distinguishable and at others not. This work cannot possibly hope to do justice to the complexity and history of the concept of causation  however the aim is to give a general overview of the concept so that we have a working understanding when we go onto consider the case law on manslaughter.

    The first principle that the law requires of the facts is that an act will only be considered the cause of an event if it is a ‘causa sine qua non’ . In other words this is a test that requires the cause to be material; ‘it must be established that the consequence would not have occurred as and when it did not ‘but for’ the accused’s conduct’ . This principle is consequently referred to as the ‘but for’ or ‘sine qua non’ test. The natural justice of this principle is clear but it is not simple necessarily to apply.

    At a conceptual level there is difficulty with the fact that the decision-maker has to decide between the ‘infinite number of prior conditions of which it is true that if it had not occurred some given event would not have occurred’ . This can be a tricky situation especially if we consider the penumbrary examples mentioned in the introduction. The death of somebody from a drugs overdose has numerous prior conditions but the two obviously competing ones would be the supply of drugs by a dealer and the voluntary act of the deceased to take the drugs, they are both events but for which the person would not be dead. The choice of these prior conditions as a ‘cause’ of the death is artificial, in reality both and many more are the cause including the friends the person made, family life, wealth, mental health and attitude to risk-taking. This is why the sine qua non test is considered to be the purely factual element of causation because once a decision-maker goes beyond these considerations they begin to make legal and policy judgements about which causes merit moral opprobrium. Hart & Honore commented on this duality within causation:

‘These two aspects of causation, even if they cannot be crudely opposed as ‘factual’ and ‘non-factual’, are of very different character. They occasion different kinds of doubts and difficulties, and different kinds of criteria are used in their resolution’

    The classic example of the application of the sine qua non test is R v. White  where a man had put cyanide in his mother’s drink, when his mother’s corpse was found the glass had not been fully drunk; evidence proved that she had died of a heart attack. The accused was acquitted of murder because his mother’s death was not contingent upon the cyanide drink and would have occurred in any case. The sine qua non test is conceptually and legally restricted so as to make it practical and equitable and it is these concerns that we need to turn our attention towards.

    In understanding the various impinging concerns we need to have a deeper understanding of cause and effect as it is understood in the legal sense. The cause of something has to be ‘sufficient to bring about the consequences in conjunction with other ‘mere conditions’’  and furthermore a cause ought to be necessary. However the ‘condition sine qua non’ doesn’t always embody these ideals and we can start to get into difficulty quickly if we do not pay attention to the fact that there are many situations where the test is satisfied but ‘the connection between X and Y has nothing to do with any form of causal connection even in the broadest sense of causal’ . The obvious example given by Hart & Honore is that it is analytically true to state that a person would not be a widow if they had not been married but there is no factual connection between the marriage and her status as a widow. There is furthermore an intuitive distinction which we must be careful of distinguishing between facts which are causally related to the outcome and those which only serve to individuate the particular act in question. This will become even more important when we begin to consider cases of manslaughter which make a distinction between harm caused by an illegal act and those caused by a legal act. There is a tendency in dealing with causation issues here to blur the issues and argue that there must be a causal link between the wrongful act and the harm caused but in a situation where a person who is driving without a licence hits someone and injures them the fact of not holding a licence is merely incidental to the harm caused.

    One final conceptual issue to remember is that most writers  present the sine qua non test as the sole factual issue of causation, this doctrine is however not the only factual doctrine and there are allowances for other approaches. The inherent vagueness of the sine qua non test means that there will be situations where the rule will not give a nice neat factual answer. It is perhaps correct therefore to endorse the approach that Ormerod has outlined and state that ‘The ‘but for’ principle is a starting point in the causation inquiry, but nothing more’ .

    As I mentioned above the duality of causation requires us to give consideration to the more complex and inherently more difficult legal side of causation. The first legal causation issue is that of culpability. The case of R v. Dalloway  is the landmark case on this issue where it was established that if the culpable behaviour of the accused in no way contributed to the result then the chain of causation would be broken. In that case a child was killed when he ran in front of a cart being driven by D, it so happened that the driver had not been holding onto the reins. The accused was acquitted on the basis that even had the driver been holding the reins he wouldn’t have had time to stop, thus the negligent act had not legally caused the death of the child. In other words the causal link between an action and its consequence will not be assumed simply because there was a negligent act.

    The culpable act must have a degree of substantiality as a cause of death, the determination of this is for the judiciary and it certainly seems to be construed narrowly as was stated by Goff LJ in R v. Pagett  ‘the accused’s act need not be the sole cause, or even the main, of the victim’s death, it being enough that his act contributed significantly to that result’. The distinction between a minimal contribution and a significant contribution has no technical meaning and in most violent criminal cases it will be what the ‘twelve men and women sitting as a jury in the jury box would regard in a common-sense way as the cause’ . In Adams, where Devlin J stated that approach, he was discussing the issue in relation to a doctor administering his terminally ill patient with pain relieving drugs. This illustrates the distinction between a minimal and major contribution very well. If the administration of drugs was considered part of a doctors duty which had the incidental effect of shortening life then this was not murder however if it was considered to be an intentional attempt to end the life of the patient then it was a substantially intervening fact so as to be considered the cause of death.

    Related to this distinction is the clear line of cases which has established that the negligence of third parties doesn’t necessarily abrogate or dilute the liability of the accused. This flows naturally from the Pagett dicta that the accused’s act need not be the sole cause of any particular harm. In R v. Benge  there was evidence that there was negligence over and above the accused’s which had contributed to a train accident. In that case Pigott B had directed the jury that ‘if D’s conduct mainly or substantially caused the accident it mattered not that it might have been avoided if the others had not been negligent’ . The flipside of this approach is what is known as the Novus Actus Interveniens doctrine which is where the act of a third party is ‘significant enough to acquire causative potency for itself’ . The border between the intervention of third parties being significant or not is sometimes incredibly difficult to demarcate to any great extent but the effect of an intervening act is to break the chain of causation.

    The easiest to identify as a novus actus interveniens is an ‘Act of God’ which is understood to have to be something occurring in nature which is random and unpredictable. There are various hypothetical situations given in the literature but the distinction seems clear; if the natural event is foreseeable then it will not be considered to be an act of god:

‘the phrase is usually used in context of very extreme climatic conditions, it can also be used to cover the equally unpredictable and unforeseeable operation of animate forces, such as the rate which gnawed a hole in the cistern in Carstairs v Taylor (1870) LR 6 Ex 217’

    The examples given in the literature show this distinction for example, if a man is assaulted and knocked unconscious then the situation can make a big difference. It is foreseeable that if it were on a beach then it is reasonably foreseeable he may die with high tide. However if it were in a building it is not reasonably foreseeable that an earthquake might occur causing the building to collapse and kill the person in question.

    Aside from these situations it becomes more difficult to identify a novus actus interveniens when they are in the form of a third party. Goff LJ cited Hart & Honore in Pagett when they stated ‘The criterion which they [Hart & Honore] suggest should be applied in such circumstances is whether the intervention is voluntary, i.e. whether it is "free, deliberate and informed’. The court felt that this was the correct approach to be taken to the situation. This automatically rules out innocent agents being forced by third parties, involuntary actions caused by a shock  and acts done by ‘a police officer in the execution of his duty acts to prevent a crime, or to apprehend a person suspected of a crime’ . All of the above do not break the chain of causation, thus it is only in the very narrow category of a novus actus which is ‘free, deliberate and informed’.

    The cases on this issue cover a very wide area and discussion of the issues raised in those cases could form a piece of work on their own right. However a brief outline of the issues raised is necessary; this work is primarily concerned with whether the chain of causation can ever be broken in cases of manslaughter. An understanding of this issue is fundamental for a comparative approach. I will talk briefly about the various categories that have considered the intervention of third parties:

•    Negligent Medical Treatment ~ The issue of negligent medical treatment is one that has occupied the courts attention on a number of occasions. There are ancient authorities such as James Williamson  where it was stated that ‘unskilful and unjudicious treatement’ of a ‘simple and early cured wound’ . This was followed in modern times in the seminal case of R v. Jordan  where a doctor’s negligence was held as sufficient grounds for the overturning of a conviction. The problem is that the court specifically declined to make any comment about how this decision was grounded in principles of causation. Furthermore in the case of R v. Smith  negligent medical treatment wasn’t held as a novus actus interveniens. The distinction between the authorities is ambiguous .  The most recent authority on the issue is the case of R v. Cheshire  where the court seems to have significantly narrowed the operation of this doctrine, here the deceased was shot in the leg and suffered ensuing respiratory problems. This required a tracheotomy to be performed. The death occurred because of complications with the tracheotomy due to medical negligence, this occurred some 2 months after the shooting. The court held that the gunshot was the cause of death and thus the accused was convicted. The rationale of the court was based on the fact that only the most extraordinary and unusual case would suffice and they argued ‘treatment which falls short of the standard expected of the competent medical practitioner is unfortunately only too frequent in human experience for it to be considered abnormal’ . This is more in line with the authorities such as R v. Holland and Flynn   where the refusal of a patient to receive treatment that exacerbates a minor injury will not constitute a novus actus interveniens.

•    Thin Skull Rule ~ It is clear from the authorities that where the victim has some pre-existent condition then this will not excuse the accused:

‘It would never do for it to go forth from this court that house-breakers or robbers…should be entitled to lay violent hands on very old or very sick or very young people, and, if their victim died as a result, to turn around and say that they would never have died if they had not been very weak or very old’  

This is a fundamental rule and obviously will not count as a novus actus.

    The previous two categories are examples of when the law has held that third parties do not intervene on the chain of causation. This is indicative of an extremely strict approach to issues of causation, they do not allow potentially large supervening factors to get in the away and if you were cynical then it would be tempting to state that consequences simply cannot be too remote from the culpable act. Only in a situation such as R v. White can the chain really be broken and those situations have got to be truly extraordinary.

Current Case Law on Manslaughter & Chain of Causation

    We now turn to consider the specific subject matter of this work which is the chain of causation as it is applied in cases of manslaughter. The major issue in this section will be focused on the penumbrary issues of manslaughter. Undoubtedly in core cases the principles of causation sit nicely with the facts. The major issue of interest are areas such as prosecution of drug suppliers where the principles of causation are stretched to their limit and open to no small amount of criticism.

    The first issue is just to briefly recap what exactly the crime of manslaughter is and how it is distinct from murder. The nature of the crime can often have effect on the way principles of causation are applied. Manslaughter is an offence which covers a vast area only ‘limited in scope…by murder at one extreme and accidental killings at the other’ . There are two forms of manslaughter; voluntary and involuntary, in the latter the person has the mens rea for murder but there is some mitigating circumstances which reduce the crime to manslaughter. Voluntary manslaughter was put into statutory form by the Homicide Act 1957 and now only occurs in three defined situations: provocation, diminished responsibility and a suicide pact. Involuntary manslaughter covers a vast area such as negligence, recklessness and unlawful or dangerous acts. This means that the scope of the crime is far from distinct; issues of causation are thus inevitably raised in cases of involuntary manslaughter where the thing that caused death may have a remote link to the accused in the particular case. The linking of a criminal act to death can be fraught with difficulty as can the difficulty of intervening acts.

    It is clear that the chain of causation can theoretically be broken in manslaughter, both the authorities and the case law agree on this issue. However, there have been a series of cases over the 30 years regarding manslaughter which seem to imply that the chain of causation may not be able to be broken even by the voluntary intervention of third parties; the chain of causation may not be possible to break.

Unlawful Act Manslaughter

The main issue here is the causal relationship between drug suppliers and their victims in the form of the recipients who overdose. The starting point for a discussion of this issue is the now seminal case of R v. Cato . The case law that has flowed from the decision in that case has shown that causation is uniquely sensitive to the particular facts of a case. As we shall see the decisions that have followed on from Cato have been severely criticised by many commentators.

    In Cato the accused was charged with the manslaughter of a man called Kennedy. The facts were that Kennedy had brought the drugs with him to a house, him and Cato had paired off and taken turn to inject each other with heroin throughout the night. Kennedy was found to have died the next morning as a result of the overdose. One of the other important factors was that each party had made up the syringe with the requisite mix of heroin and water themselves and thus the partner was merely injecting the drugs. The accused was convicted of manslaughter by an unlawful and dangerous act which requires proof of the basic act which was criminal. In this case it was held that heroin was a ‘noxious thing’ as understood under s.23 of the Offences Against the Person Act 1861 and that this base crime was sufficient to establish manslaughter. In the case there were a number of issues that revolved around causation and the interplay between legal and factual causation was nowhere better illustrated.

    In this case causation in the strictest sense was not really an issue. There was an attempted argument on appeal that it had not been proved satisfactorily that it was heroin that had caused death but on the evidence this was rejected. The really interesting issues were sidestepped in this case because per se under the Misuse of Drugs Act 1971 the actions of Cato was an offence. The dubious judicial chicanery of using s.23 of the Offences Against the Person Act 1861  aside the question sidestepped the issue of whether consent would have broken the chain of causation in a charge of manslaughter by gross negligence or recklessness, Lord Widgery in Cato specifically stated that the argument was not pertinent, he went as far as to say that they would have used unlawful possession as a substitute to s.23. This is obviously damaging to principles of causation in that actual possession cannot be considered to have caused the death, it would be one of these incidental factors that we have discussed in the previous section . Furthermore the case law under the Misuse of Drugs Act 1971 precludes any alternative such as supply of drugs interpretation: R v. Maginnis  and R v. Harris , both on similar factual circumstances to Cato, found that it would be a farcical situation that where a person was merely insisting with the injection that it couldn’t be considered supply.

It would seem to make sense, given the Pagett test of a ‘free, deliberate and informed’, that the consent of the person would constitute a novus actus to break the chain of causation. However, this issue of the supply or administration of drugs has come before the courts on numerous occasions since Cato and we can derive a number of lessons from these cases about the operation of causation in cases of manslaughter however it has to be argued that the logic in Cato ‘seems to stretch credulity’ .

    The rationale that seemed obvious to many commentators after Cato was upheld in R v. Dalby  . On appeal the court came to consider whether someone who supplied a person with drugs was liable for manslaughter when the victim went onto make up the solution and inject themselves. The court decided that ‘Where the charge of manslaughter is based on an unlawful and dangerous act, it must be an act directed at the victim and likely to cause immediate injury’ . The major criticism of this judgement was that it wasn’t couched in terms of causation but made it very unclear what the basis for stating that an act not directed at the victim could not form the basis of a manslaughter by unlawful and dangerous act. Common sense seems to imply that it is the lack of causal relationship but this was not endorsed by the court. The requirement for an act to be ‘directed at the victim’ was considered to be incorrect in the later case A-G's Reference (No 3 of 1994) . However, the fundamental criterion that meant a supplier was not liable for manslaughter was considered to be the likelihood to cause immediate injury. This makes it a causational issue.
    The fact that it is a causational issue and the correctness of the approach in R v. Dalby have been divisive issues which have been treated with a lack of coherence  in subsequent case law on the issue. The case of R v. Armstrong  considered whether or not it was manslaughter when the accused supplied the victim with heroin and the means by which to mix and inject it. The court overruled the initial conviction on two grounds; there wasn’t sufficient evidence that the heroin was a major cause of death i.e. it fell foul of the de minimis principle and more importantly that the voluntary act of taking the drugs constituted a novus actus interveniens and consequently broke the chain of causation. Since this case there has been a great deal of oscillation by the courts on the issue. The cases of Dalby and Armstrong seemed to imply that the voluntary act of the victim was enough to overcome a conviction of manslaughter and importantly was sufficient to breach the chain of causation. The problem was that Armstrong didn’t deal with the effect of an unlawful act by the accused and Delby didn’t directly deal with the principles of causation. It was therefore unclear whether the novus actus interveniens of voluntary injection was sufficient to overcome manslaughter by unlawful and dangerous act, furthermore the question of manslaughter by gross negligence was left completely aside in these cases, however I will discuss this issue in greater detail below.

    The next case to consider the issues was the much maligned R v. Kennedy  which came to consider a different factual circumstance. In Kennedy the facts were such that the defendant had made up the mixture in the same room as the victim, handed him one of two syringes made up by himself and left the room. Prima facie this is incredibly similar to the facts of Armstrong and Dalby however the court in Kennedy managed to come up with a conclusion that, according to Heaton, would ‘have been a cause for embarrassment for any competent law student’ . Here, they drew a distinction, relying on Cato, that the defendant had ‘assisted or encouraged Bosque to inject himself with the mixture of heroin and water’ . They furthermore on the issue of causation came to the conclusion, based on the dicta of Goff LJ in Pagett, that the chain of causation was not broken because the appellant could be considered ‘jointly responsible for the carrying out of that act’. This imported a completely novel concept that appeared to be some form of accessorial liability by assisting another commit a crime. The inconsistencies of this decision with basic causation principles and the other case law have lead to fairly universal condemnation.

Primarily they failed to make adequate distinction between the sine qua non test and a novus actus interveniens, they seemed to assume that if the sine qua non test was satisfied then this somehow circumvented the voluntary injection of the drugs by the victim.  The legal argumentation was also completely flawed in that case because the victim hadn’t in himself been committing a crime so the defendant couldn’t have been art and part guilty of assisting him and even more forcefully a person cannot commit manslaughter on himself so the accused cannot be accessorially liable for that manslaughter. Furthermore, the rationale of dangerousness in Kennedy and the distinction from Dalby that the court made are not supported when we consider the dicta in the latter case that we mentioned above about ‘immediate harm’. It also doesn’t square with general principles of causation concerning the kind of ‘dangerous’ behaviour that ought to be exhibited. In the cases of R v. Watson  and R v. Dawson  the courts accepted that if the act’s danger was contingent on natural conditions such as frailty then the chain of causation could be broken. Thus in both cases there were robberies, in Watson the victim was an old man but in Dawson the victim was seemingly healthy. In both cases the victims suffered heart attacks but the courts held the dangerousness of the act was only established in Watson because the harm was foreseeable. It is at least arguable when we consider the fact that given Heroine can be a safe drug when taken in the correct quantities and with knowledge of the particular person’s tolerance level that there would have been no way the defendant’s actions could be considered dangerous. This aside the court in Kennedy made no reference to Armstrong, however it might be possible to cynically distinguish the two on the basis that the accused in Armstrong couldn’t be bothered to make up the mixture and left it to the victim however in Kennedy he was the one who did the preparation.

Kennedy’s rationale was never very clear and more recent judgements have even further complicated the issues by attempting to distance themselves from the dicta of that case whilst still achieving the same result. It seemed that lucidity had been achieved when the decision in R v. Dias  was handed down by the court of appeal; on facts almost identical to Kennedy the court quashed the conviction of manslaughter. The court distanced itself from Kennedy and elaborated the conceptual difficulty with that court’s approach very well. Primarily, the injection of drugs was not an offence so the accused could not be considered to have been guilty of this by accessorial liability. Separately, the issue of causation cannot be made out as between the illegal supply of drugs and the death of the accused because of the voluntary actions of the victim. And finally the main crux of confusion in Kennedy was highlighted from this point:

‘It follows from that that the appellant could only have been guilty of manslaughter as a secondary party and not as a principal. But in that case who is the principal guilty of manslaughter? As there is no offence of self-manslaughter, it is difficult to see how the appellant could be guilty of that offence as a secondary party because of his encouragement or assistance to Escott over the injection of the drug’

    The decision in Dias amply highlighted the difficulties the courts had been struggling with because of conflation of causation with the unlawful act itself. The failure to realise that accessorial liability to manslaughter where the principal was the victim made no logical sense was at the heart of this confusion. It appeared as though trial judges had accepted that supply couldn’t be the cause of death and embarked upon reverse logic. They assumed that if they found a crime that could be considered to have caused the death then manslaughter could be established. This is to make a fundamental mistake about manslaughter by unlawful and dangerous act, the first step is to establish that an unlawful act has occurred and then to see whether that act caused the death. Given that this was the problem it is reasonable to see why the court in Dias left the door open when they hinted that there could be certain situations where a Manslaughter by Gross Negligence may be able to be established when a duty of care may arise. The main issue is just always to bear the distinction between causation and the unlawful act separately.

    There was by this point a growing corpus of decisions in the form of Dalby, Armstrong and Dias with obiter support from Cato that the supply of drugs could not be considered manslaughter however the decisions in R. v Finlay  and R. v Rogers  have ‘muddied the waters’ . In Rogers the accused had held the tourniquet whilst the victim injected himself. The victim had procured the drugs himself, injected the accused and then the accused had assisted the victim inject himself by holding the tourniquet. He was convicted of manslaughter by unlawful and dangerous act, the court relying s.23 of the Offences Against the Person Act 1861. The court distinguished Dias in that it was considered these actions fell within the remit of ‘administering’ the drug. Like in Cato the court were obviously using a real stretch of logic. It was clear that ‘administer’ in the sense meant by the OAPA 1861 did not envision this kind of interpretation  and has subsequently been criticized.

    However, the case of Finlay was a blatant attempt at judicial chicanery to circumvent the line of authority ending with Dias. The facts were identical to Kennedy and Dias but the result was different. They re-established the chain of causation between the supplier of drugs and the victim by relying on s.23 of the OAPA 1861 where it states ‘Whosoever shall unlawfully and maliciously administer to or cause to be administered…’ The court used the famous case of Environment Agency v Empress Car Company  to establish that in fact the chain of causation could not be broken by the voluntary act of the victim as a novus actus interveniens when that action would be an ‘ordinary occurrence’ as opposed to an ‘extraordinary’ occurrence . The court in Finlay felt that the voluntary injection was an ordinary event and would not break the chain of causation, thus the voluntary nature was just a factor to be taken into consideration. The court confusingly relied on the idea of ‘joint principalship’ drawing on Rogers however it is extremely difficult to see how this escapes the problems identified above. The victim and the accused cannot be joint principals in an s.23 offence and also cannot be joint principals in manslaughter. In both of these cases there is a requirement of a third party. Furthermore the use of Empress Cars as authority is extremely dubious given that the case is also one which is subject to a high degree of criticism. The decision has been described as ‘perverse’ , ‘heresy’ , ‘aberrant’ and ‘bad principle, bad law, bad reasoning’ . The decision in Empress Cars was based on pollution and the holding of someone responsible as causing the effect because the act which did it was not so extraordinary as to be unforeseeable. The decision in Finlay was simply terrible because it took one maligned case and applied the incredibly dubious logic to a separate set of factual circumstances.

    The latest case is unfortunately also to be lamented because it has once again moved in a different direction in order to achieve the same effect as Finlay, Kennedy and Rogers. This involved the second appeal of Kennedy to the court of appeal . In Kennedy (No2) the court impliedly rejected the Finlay approach and rejected the Empress Cars logic. Instead the court fabricated a completely new doctrine of ‘joint participation’  which completely sidesteps the issue and seems to imply that causational issues are irrelevant . The court argued that in situations such as Kennedy the accused can be considered to be acting in concert with the victim to commit an offence under s.23 OAPA 1861. This seems prima facie to ignore the fact that a s.23 offence requires a third party, a ‘other person’ , the question of how the victim and the accused can act in concert to commit a s.23 OAPA 1861 is tautologous and amounts to the same argument in Kennedy. The whole decision tries to sidestep causation. The court is fully aware if it can establish an s.23 OAPA 1861 then it can establish causation in the natural manner. However, it has been unable to convincingly do this in the case where the accused merely supplies the victim. The reliance on joint participation cannot resolve the problems created by the drug supply question for two very important reasons that Ormerod & Fortston identify:

   Primarily, the court in Kennedy (No2) recognised an accused cannot be held liable as a secondary party for the reasons explained in Dias. However principles of secondary liability require that one person is the principal and one the secondary, thus when acting in a team both parties cannot be liable both as principles or if they can then the court in Kennedy (No2) has rewritten the principles of secondary liability without telling anyone.

    Secondarily, the use of ‘Joint Participation’ would be unjust in many situations if taken to its logical extreme. Ormerod & Fortston give a specific example here:

D who supplies the bullets for P's gun for him to murder V now becomes a murderer as a principal, not as an accessory. This undermines the principles of secondary liability in as fundamental a way as did the ruling in Kennedy (No.1)’

    The clear message from the case law is that the judiciary seem willing to convict at any cost no matter the damage to principles of causation and the general criminal law. The application of the law in these cases is incoherent and from Kennedy (No1) to Kennedy (No2) we have heard four different justifications all completely flawed. The chain of causation in these cases is continually held not to be severed by the seeming novus actus interveniens of voluntary injection but the rationale for this is very unclear.

    It is clear however from other case law on unlawful act manslaughter that the drug cases are not consonant with other principles of causation. In R v. Williams  the court talked about the voluntary act by a third party on the chain of causation and stated that the voluntary act had to be ‘within the ambit of reasonableness and not so daft as to make it his own voluntary act which amounted to a novus actus interveniens and consequently broke the chain of causation’. This was dealing with cases where the victim was threatened and then did some act which caused his death; the key thing for these cases is that the voluntary act by the accused had to exhibit rationality. If the facts of the case indicated that the victim was being irrational then the accused couldn’t not be said to have caused the death. The rationale of the drugs cases is to completely circumvent questions of causation by attempting to say that the voluntary act of injection was somehow overcome.

    The principle of novus actus interveniens when the victim or another person makes a ‘free, deliberate and informed’ decision is one of the most important concepts of causation. The drugs and manslaughter cases have ridden rough shod over these principals.

Gross Negligence Manslaughter

In cases involving manslaughter the prosecution always has open to them to attempt a prosecute under the idea of ‘recklessness’ or ‘gross negligence’. The issues of causation are also pertinent here and given the speculative nature of the crime the chain of causation can sometimes be difficult to establish. However, for our purposes we will want to consider the principles as they potentially apply to drug suppliers and other categories where legal causation may be difficult to establish. There was considerable confusion in recent years but the confusion seems to have been settled by the court in R v. Adomako . Following that case it is clear the prosecution must establish three criteria:
•    Must be in breach of the duty of care under established principles of negligence.
•    The negligence must have caused death.
•    The action must be considered as ‘gross negligence’

In the drug supply cases it was suggested by R v. Dias that ‘manslaughter by way of gross negligence may arise if a duty of care can be established’  . Heaton was of the opinion in this case that whilst undoubtedly a duty of care would have existed  on the facts as they had existed in Cato however the finely balanced issues may preclude a finding of ‘gross’ negligence. In cases such as Dalby, Kennedy and Finlay where the victim self-injects it is hard to see how this progresses the issue of causation any further.

The main problem with drugs cases is that the reasoning is completely circular and very difficult to grasp without detailed contemplation of potential analogies. The crucial distinction for the whole case law is between the situation in Cato and those in Kennedy. The difference between those two is that Cato could be considered to have ‘caused’ the death of the victim in that he was the one doing the injecting, the consent of a victim is no defence for good reason. Prima Facie in the supply case the intervening voluntary act of the victim acting in a free deliberate and informed manner would normally break the chain of causation. What the court did in Kennedy (No2) was to leave the law in confusion, especially because they didn’t couch their approach in terms of causation. The problem was amply highlighted by Heaton:

‘An individual, D, might be regarded as a factual cause of the prohibited result where his action combined with others brings about that result. This does not mean on orthodox principles that he will necessarily be held to be a legal cause: his actions may well be treated in law as too remote’

The use of joint participation to establish the s.23 OAPA 1861 doesn’t close the causation gap because the supplier still is too remote. The court attempted to circumvent this inevitability by relying on R v. Latif and others :

‘The free, deliberate and informed intervention of a second person, who intends to exploit the situation created by the first, but is not acting in concert with him is held to relieve the first actor of criminal responsibility’

    Thus causation becomes reliant on the perception of ‘acting in concert’ however this causes a beehive of problems. Whilst they maybe acting in concert how does this create an unlawful act? S.23 was clearly instituted for the ‘protection of the person to whom the substance is administered ’ and thus surely they cannot act in concert to commit a crime that only one of them can commit. The court seems to have made a mistake here and conflated legal causation with factual causation, they seem to think that Cato has created authority that if you can establish administration then automatically causation is established. It is clear that this is not the case however the court in Kennedy wants to shift the legal basis the facts remain the same there was not a traditional causational link.

The basic answer to whether the chain of causation can be broken in cases of manslaughter is that theoretically and it has been broken in numerous cases. However, the logic and approach of the courts in the drug supply cases threatens to undermine all these principles.  

2: Comparison to other areas of Criminal Law

The decisions in Empress Cars and the drug supply cases have been heavily criticised as having caused ‘a revolution within the field of criminal causation’ . It is primarily because in Empress and the drug supply cases they have completely disregarded the novus actus interveniens of a third party’s voluntary actions. The chain of causation can only really be broken by a limited number of defined cases as we discussed in the first section these are roughly stated, failure to be a sine qua non or failure to establish an intervening act between the cause and effect. The removal of the voluntary action intervention distorts these principles and if taken to its logical extreme could cause serious defects in criminal causation. However, the principles of causation, as traditionally understood always seem to find the ‘best illustrations supporting the thesis…in cases other than homicide’ , this is undoubtedly an accurate description.

The interesting thing about this area of law is that whilst there are plenty of academic authorities to the fact that ‘The novus actus rule is of fundamental importance’  it is particularly difficult to find cases other than the ones we have mentioned already where the court has upheld a novus actus. However, one of the main justifications for the existence of the rules on causation is the distinction in criminal law between principal and accessory ; it is in these cases that we are able to find dicta which are at odds with the approach in the drugs supply cases. The whole basis of accessorial liability is based on the fact that the accused didn’t in fact cause the harm, in fact he may have had absolutely no intention to commit the crime but he can still be an accessory to the crime.

Thus take the famous case of R v. Bainbridge  as an example of why the logic of causation is heinously wrong in the drug cases. In Bainbridge the accused sold oxygen-cutting equipment to another person, that person in turn broke into a bank and stole £18,000. The accused was convicted as an accessory to the theft. Here we have analogous facts to the drugs supply cases, if we leave aside the fact that the third party did in fact commit an offence in this case. One cannot be accessory to theft where there are only two parties. For example if the third party had taken the oxygen-cutting machine and broken into his own house and ‘stolen’ £18,000 then no offence is committed by the accused. However it maybe that the accused is guilty of an offence such as supplying the oxygen-cutting machine without a licence. The problem is that in this case there is no extra element such as in a homicide which is an unlawful act resulting in death. Therefore lets say when the third party is using the oxygen-cutting machine breaking into his own house he cuts himself, there is no crime of self-injury but there is one of assault. Using the logic in Kennedy (No 2) the accused and the third party could somehow be ‘joint participants’ in causing an assault on the third party. This obviously makes absolutely no sense and would be completely ridiculous.

Thus the approach taken in the manslaughter cases doesn’t sit easily with principles of accessorial liability because as I have argued above legal causation seems to have taken precedence over factual causation which means that establishment of who is the principal and the accessory become questions of legal construction as well as fact.

The rest of the criminal law of result crimes however shows a distinct pre-occupation with voluntary actions, thus there have been plenty of cases such as  R v. Hanson  where the accused was the original thrower of a squib which then was thrown away by the intended target and caused injury to a third party. The fact that the throwing of the target was not voluntary meant that it couldn’t break the chain of causation. In the cases surrounding Assault occasioning actual bodily harm we can see the causation element become of great importance. In R v. Roberts  the assault was an attempt to remove a woman’s coat which occasioned her to jump out of a car and sustain injury. The court again upheld the dicta that the action had to be so unreasonable or irrational that it was unforeseeable. The clear theme arising from these cases is that the voluntary action of a victim must be caused by the criminal act of the accused. This can be established in these ‘escape’ cases such as Roberts because the facts bear out that it was the cause, however it is clear in drug supply questions the implied assertion and the one accepted in Dias is that the supplier cannot be said to cause the death. The foreseeability issue that Finlay raised was taken in the wrong context, it is clear the stressed mindset of the victim was what made the harm foreseeable. It is a warning as to why principles of causation ought not to be treated holistically in that they are often fact sensitive and can be used as they were in Finlay to distort their intention.

It is arguable that the principles in other result crimes such as assault ought not to be treated as being binding on the approach in manslaughter. We will see that the crime can make a difference to the way causation is established in the next section when we turn to look at the approach in other jurisdictions.  

3:Comparisons to other Jurisdictions

The issue of the supply of drugs has been raised in a few cases north of the border and it is interesting to look at their approach as a comparator. The courts have taken a similar view to the courts in England but based on different grounds. However, the decisions north of the border have consistently found suppliers to be guilty and that the Novus Actus Interveniens of voluntary behaviour doesn’t break the chain of causation. In both Khaliq v. HMA and Ulhaq v. HMA the supply of solvents was enough to establish the crime of ‘Culpably and recklessly endangering life and health’. In Lord Advocate’s Reference (No1 of 1994)  the court considered a directly analogous case to those of the drug supply when the accused was charged with the Scots Law equivalent of manslaughter, Culpable Homicide. On Appeal to the High Court after an initial acquittal the High Court upheld the views of Khaliq and Ulhaq and specifically stated:

‘The purpose and the intention of the supply libelled was that the solvents should be abused, employing the means of abuse provided. The solvents were, it is averred, abused, and this could be regarded as the expected, intended and probable consequence of the supply’

This knowledge of abuse thus in Scots Law was held to be vital because the foreseeability of the use meant that the chain of causation was not broken. They argued that the distinction between supply and administration was so negligible where ‘supply is closely related to the use by words of instigation or by some act of instigation on the part of the panel’ . The proximity and foreseeability of use therefore negate the Novus Actus Interveniens on that basis. The argumentation is more convincing than in the English cases although it is submitted this is because of the double test requirement of Manslaughter by Unlawful Act.

United States of America

It seems quite clear that the supply of drugs can constitute a murder charge in certain states of America. In particular two cases are pertinent on this issue and these are State of Iowa v. Albertson  and Lofthouse v. Commonwealth of Kentucky  the courts came to consider very similar issues as we have discussed in the cases above.  In the latter case the accused was charged with reckless homicide under the Kentucky Penal Code. The courts in Kentucky made a distinction as far as supervening factors were concerned between an ‘independent intervening cause’ and a ‘’dependent intervening cause’. The latter did not relieve the accused of criminal responsibility unless the human act was completely abnormal, the former would relieve unless it was ‘reasonably foreseeable at the time of his conduct’ . The issue of causation was thus circumvented on the basis of foreseeability based on the inherent dangerousness of the drugs:

‘The Commonwealth posits that, as a matter of law, the act of furnishing narcotic drugs to another creates a substantial risk of death to the transferee sufficient to convict of either second-degree manslaughter (awareness of and conscious disregard of the risk) or reckless homicide (failure to perceive the risk)’

The test is very similar to the Empress and Finlay approach and the distinction between ordinary and extraordinary occurrence that was made in the latter case, however the main issue here is that whilst foreseeability is the test they have specifically attuned the offence so that the issue of causation is correctly centred on the foreseeability of the harm resulting from an action. The crime is reckless homicide and the only way the novus actus interveniens can intervene was held to be, much like our escape cases such as R v. Roberts  and R v. Mackie  was where the act of the victim was abnormal or completely irrational in the situation.

Interestingly the US has completely rejected what was known as the felony murder rule which is defined as:

‘If death ensues as a consequence of the commission of a dangerous felony,
the intent to commit the dangerous felony provides the element of intent necessary to convict of the homicide’

The court in Lofthouse discussed the case law surrounding this crime which bears obvious analogies to manslaughter by unlawful and dangerous act. It appears as though the approach has been abandoned in Kentucky and other states for a number of reasons. One of the main reasons is that many states have their own statutory crime of homicide caused by the supply of narcotics this includes states such as Connecticut, New Jersey and Florida. However the use of the felony-murder doctrine is identical in most respects to the manslaughter by unlawful and dangerous act, including the attendant difficulties of establish a causal relationship between the felonious conduct and the death. The approach in the seminal case of Heacock v. Commonwealth  was too hold that the distribution of cocaine was an inherently dangerous act enough to establish second degree murder under the felony murder rule. It has subsequently been held in numerous cases that the supply of drugs under the felony murder rule will constitute a crime see Ingleton v. State  , State v. Whitted  and Sheriff, Clark County v. Morris .

Australia & South Africa

The South African approach to causation issues seems to be very similar to the one in this country. In two particular cases; S v. Grotjohn  and S v. Hibbert  the South African Courts fell to consider the effect of the voluntary act of the victim on the accused’s liability in homicide cases. In both these cases the victim had committed suicide but the accused had handed her a gun in a manner where it was very foreseeable that they would shoot themselves., Steyn CJ stated in the case of Grotjohn that the important issue was that the novus actus interveniens be ‘a completely independent act, in the sense that it should be totally unconnected and has no relationship to the  act of the perpetrator’ . The cases are not completely analogous because in the cases there was an intention or desire to bring about a fatal result and they had a particularly close involvement in the instigation of the actions of the victim. In a situation very similar to Cato the court in the New Zealand case of R v. Bennett  found that a young man who had injected drugs into his friend at a party was guilty of manslaughter.

Continental Approach to Causation

There is a scarcity of directly comparable cases on causation in drug supply cases however Hart & Honore have written about the continental attitude to causation, particularly as it applies in Germany. This is not the place for a dissertation on the theory of causation on the continent but it is premised on a theory of conditions, which treats a particular consequence to be made up by a series of conditions which are then tested to see whether they are a necessary condition by hypothetically removing certain conditions and substituting others . It seems that the vast majority of continental jurisdictions, such as Germany and Spain, take this approach to causation and in relation to voluntary interventions there law seems to be consonant with the principles that we supposedly adhere to. In Spain the chain of causation is broken where there is ‘express and deliberate fault of the victim’  which in the drug supply cases unfortunately there is a paucity of accessible Spanish decisions to see whether this would hold in the drug supply cases.The German courts in a number of cases have rejected the use of the voluntary intervention  of a third party as a Novus Actus Interveniens, one particularly analogous decision was based on the liability for negligent manslaughter where the accused gave poison to a third party and the latter used this to kill his victim . In another German case it appears as though there was an attempt to prosecute an arresting officer for the death of the victim when he was shot whilst in custody by a separate officer. The court stated that the act has to have ‘the tendency in the light of experience to bring about the consequence of the kind in question’ . In another case a man who abandoned his daughter who was pregnant was held to be liable for the negligent manslaughter of the child when the girl committed suicide .


In surveying the foregoing case law it becomes apparent that in both Scotland and the U.S.A the law is in effect much the same as the result in Kennedy (No 2). However, what we see is that both on the continent and in these other jurisdictions is that undoubtedly any liability is based upon the idea of negligent manslaughter or culpable homicide because death was reasonably foreseeable. In all the jurisdictions the interrelation between foreseeability and causation is that the causal link can be established by negligence. This is because in those situations the seemingly voluntary action of the victim is foreseeable, or ought to be, so the accused is deemed to have been aware of the dangers and continued any way. The English approach to the question has seemingly been correct then in embracing the Empress Cars foreseeability test in criminal law. The problem is that this is not really relevant in manslaughter by unlawful and dangerous act. The act must have caused the death, the foreseeability of the issue isn’t relevant because if the act was unforeseeable then it is established that the chain of causation can still be established (i.e. the person had a thin skull or a rare disorder). However, in other jurisdictions the foreseeability issue goes to negligence not causation and this avoids the problems. In all jurisdictions the supply of drugs is a causa sine qua non however the chain of causation is held not to be broken because the crime is consequent upon negligence by definition whereas the manslaughter cases require a standard causational link between the act and death.

    This work has concentrated predominantly on the drug supply cases in manslaughter; this mirrors the domination that is given to them in all the textbooks and articles on the issue. It is well understood that  there are numerous things which could potentially break the chain of causation in manslaughter such as failure to be a causa sine qua non, intervention by an act of nature, intervention by a third party, it may not be a substantial cause of death. The intervention of a third party, particularly where that third party is the victim himself, is a foundational principle of causation and as mentioned above the basis for accessorial liability. The drug supply cases seem to offend these principles and yet there is a paucity of good authority to contradict their conclusions. In many ways the conclusions in Kennedy (No 2) defy logical explication, however it is an obvious attempt at maintaining the policy that drug dealers can be guilty of manslaughter. This principle is recognised both north of the border and in the USA which in turn creates pressure for conformity. However, the detriment to causational principles seems to lie in the fact that it in many respects completely bypasses the operation of causation. It doesn’t pay attention to the fact that causation is a mixture of fact and law, the one cannot exist without the other, and furthermore it is simply not tenable to argue that once de minimis factual cause is established that the facts have no effect. Kennedy (No 2) tried to create some kind of legal chain of causation, whereby the establishment of ‘administration’ by constructing the facts so as to fit into the definition somehow changed the facts from what they were, thus legally all administration causes death where in fact it does.

The chain of causation can be broken in cases of manslaughter; it is clear from American authorities such as People v. Elder and Smith v. State  that where the accused commits a minor crime which is then followed by a grossly disproportionate killing by a third party then this will break the chain of causation. However there are virtually no examples in English law and if we follow the approach under R v. Dear  as we discussed above the chain of causation in cases concerning homicide may be incredibly difficult because even the most unreasonable behaviour of the victim in exacerbating an initially trivial harm are not relevant in breaking the chain of causation. It is as well to remember however that this doesn’t necessarily entail a conviction of manslaughter, causation is primarily concerned with the actus reus of a crime and the mens rea may have a very important bearing on the outcome.



  • Allen, Michael        Textbook on Criminal Law        2005 / OUP / 8th ed.
  • Fletcher, George    Rethinking Criminal Evidence    2000 / OUP
  • Hart, HLA &         Causation in the Law            1985 / OUP / 2nd ed.
    Honore, T
  • McCall Smith, R &    Scots Criminal Law            1997 / Butterworths / 2nd
    Sheldon, D                            2nd ed.
  • Ormerod, David    Smith & Hogan Criminal Law    2005 / OUP / 11th ed.


  • Heaton, Russell    Dealing In Death                2003 CLR 497
  • Principals? No Principles            2004 CLR 463
  • Ormerod, D &        Drug Suppliers As Manslaughterers (Again)    2005 CLR 819
    Fortson, R
  • Padfield, N        Clean Water And Muddy Causation: Is     1995 CLR 68
  • Causation A Question Of Law Or Fact, Or
  • Just A Way Of Allocating Blame
  • Phippen, L &        Drugs and manslaughter            2005 NLJ 1054
    Radlett, D
  • Reed, Alan        Involuntary Manslaughter And Assisting    2003 JCL 431
    Drug-Abuse Injection
  • Court of Appeal: Causation and Assisting    2005 JCL 384
    Drug-abuse Injection
  • Toczek, L    Is the supplier also the killer?                2002 NLJ 595
  • Weait, M    Taking the blame: criminal law, social        2001 JSWFL 441
     responsibility and the sexual transmission of HIV
  • Williams, R    Policy And Principle In Drugs Manslaughter Cases    2005 CLJ 66



Source: Essay UK -

Not what you're looking for?


About this resource

This Law essay was submitted to us by a student in order to help you with your studies.


No ratings yet!

Word count:

This page has approximately words.



If you use part of this page in your own work, you need to provide a citation, as follows:

Essay UK, Law Essay on Chain of Causation cases of Manslaughter . Available from: <> [22-01-19].

More information:

If you are the original author of this content and no longer wish to have it published on our website then please click on the link below to request removal:

Essay and dissertation help


Lederfarbe 177 Farben 40ml PARVA | Anil Kapoor | Maldives