Automatism as a Legal Defence

Edward Argar Excerpts

Tuesday 14th May 2019 Westminster Hall

[The Parliamentary Under-Secretary of State for Justice (Edward Argar)]

I suspect I will, Mr Owen. It is a pleasure to serve under your chairmanship.

I congratulate the hon. Member for Huddersfield (Mr Sheerman) on securing this debate on automatism as a legal defence. I am conscious of his work in this area and across a range of aspects of the operation of the law, particularly regarding the impact on victims. I know that he has taken a particular interest in a number of tragic cases. Although he will appreciate that I cannot comment on individual cases, especially where charging decisions are concerned, I know his work in the area.

Given the limited examples of the use of automatism as a defence, it might be worth elaborating a little, as other hon. Members have, on the current law and its operation. I will also seek to share the Government’s general approach to reform of the law in this area, and the Law Commission’s.

Generally, automatism is a defence to a criminal charge where the defendant’s consciousness was so impaired that he or she was acting in a state of physical involuntariness. It is more than not intending something to happen; it is not being aware of those actions. There are two distinct types of automatism. First, there is insane automatism, which stems from an internal cause or disease of the mind. Where this occurs, the proper verdict is not guilty by reason of insanity. Secondly, there is sane or non-insane automatism, which stems from a cause other than a disease of the mind, an external cause that leads to a loss of control. Where this occurs, the proper verdict is not guilty. The hon. Member for Huddersfield suggested, and I think I am quoting him accurately, that there was an increasing tendency by clever lawyers to seek to use this defence, but I have to say that there is no evidence of that. He himself mentioned the stats, and this comes back to the statistical point, which I will touch on shortly. As always, the hon. Gentleman makes his point forcefully, to highlight the issue that he believes is behind this.

[Mr Sheerman]

How do we know how many times the Crown Prosecution Service does not prosecute because it thinks it will not be able to get a conviction because of the use of automatism?

[Edward Argar]

I will come back shortly to the two tests that the shadow Solicitor General mentioned. On the statistics point, it appears that automatism is rarely used as a successful defence, and that sane automatism, which is what most people assume that to be, is extremely rare because it is very hard to prove. However, the hon. Member for Cumbernauld, Kilsyth and Kirkintilloch East (Stuart C. McDonald) also made the point about the statistics. It is a matter for the CPS, but we can take back to that service and to other agencies the question of examining whether there are better ways to identify trends and the statistical evidence bases underlying them. I do not have an answer now, but that is something we can take away and look into.

[Nick Thomas-Symonds]

This is an issue that I have raised before in other contexts. Sometimes we are told by the CPS that it has to look through the file of each case to pick up certain data, but in such cases, particularly where there are deaths, as there might be in a road traffic case, it would be useful to look at what data is collected and reviewed. I hope that the Minister will pass that on to the law officers who superintend the CPS.

[Edward Argar]

I am happy to do that, not least because in her previous ministerial role the now Solicitor General would have been taking this debate. I am sure that having prepared for it in advance, as she usually did, she will be well aware of the issue, and I am happy to raise it with her.

The distinction between the two types of automatism is important for reasons other than the verdicts. In insane automatism, the burden of proof is on the defendant to show that he or she did not know the nature of the act committed. However, in the case of non-insane automatism, the burden remains on the prosecution to satisfy a jury beyond reasonable doubt that the defendant has not had a total loss of control and is, therefore, guilty of the offence.

To illustrate that further, there is the example of incidents that occur as the result of hypo or hyperglycaemia. If a defendant argues that the act was caused by the administration of insulin leading to hypoglycaemia, that is an external factor and the defendant will be acquitted unless the prosecution can show this to be untrue. However, if the defendant argues that the incident was due to diabetes causing excessive blood sugar hyperglycaemia, that would be an internal factor and the onus would be on the defendant to prove, on the balance of probabilities supported by medical evidence, that he or she was not guilty by reason of insanity.

The cause of automatism can understandably be confusing to many. I will give an example that is possibly a reflection of how the law is interpreted and the difference between what is, in legal terms, the interpretation and what anyone else might read it as. For example, epilepsy is a disease of the mind. When an epileptic seizure results in an assault, for instance, a successful defence would be an insanity verdict. Of course, most people would not consider that an epileptic seizure amounted to insanity, but that is how the law would be interpreted in a narrow, legal context.

Acts committed while suffering from concussion, sleepwalking—which we have touched on—amnesia, and even post-traumatic stress disorder may amount to an automatism defence. As alluded to earlier, such cases rely heavily on medical evidence as to the nature and causes of the loss of control. Perhaps the most famous illustration of automatism was in the context of an example given in one of the leading cases, Hill v. Baxter in 1958—that of a driver who causes an accident after being stung in their cab by a swarm of bees.

I will cover two final points before ending my survey of the current law and moving on to the future. I know that the hon. Member for Huddersfield has a particular interest in driving offences in the context of this debate, although his interest goes wider than that. Automatism is a defence even against strict liability offences; I mention that because although that defence occurs infrequently, it is most commonly reported in relation to driving offences. Significantly, that defence applies even when the offence does not require intent, such as with dangerous driving.

As my hon. Friend the Member for Cheltenham (Alex Chalk) and others have highlighted, when automatism arises from prior fault or voluntary conduct, that usually—but not exclusively—means that the defendant was taking illegal drugs, or that alcohol was involved, for example. When a person is taking substances other than in accordance with medical direction, and the crime is one of basic intent such as an assault, the defence should fail if the substance taken is known to cause aggression or the consequences that caused the offence. It is not a defence to be completely out of it on drugs or due to alcohol, and as a result commit an offence, however unknowingly. Those questions, though, will often be left to the jury or to magistrates. It is worth stressing that the defence is much more narrowly drawn than many people might imagine.

Let me turn again to why the defence exists and briefly touch on the Scottish situation, which the hon. Member for Cumbernauld, Kilsyth and Kirkintilloch East alluded to. I understand that in Scotland the context is different, in that an automatism plea is treated as a denial of mens rea. In other words, if acting as an automaton, a person lacks the essential mental element of a crime and as a result should be acquitted of an offence. However, the Scottish courts have clarified that, for the defence of automatism to succeed, there must be a total alienation of reason that is caused by an external factor that was not self-induced or foreseeable. The internal versus external distinction therefore applies as a test in Scotland as it does in England, albeit in a slightly different form, and the defence will also fail if the defendant’s state is self-induced. Although the test of what amounts to insanity is different in Scotland from that in the jurisdiction of England and Wales, the defence—although it takes a slightly different approach—is none the less very similar in its application and consequences for defendants.

As I believe all who have spoken today have also said, I do not think that anyone would contend that a person who commits an act because of loss of control and through no fault of their own should be held liable for that offence. In such cases, it is also unlikely to be in the public interest to prosecute, as a conviction would not be secured. If there is doubt about whether the defendant contributed to their loss of control, then those questions may be tested in court. Of course, it is very difficult for innocent victims of these acts, who may themselves be injured or bereaved as a result of them, to accept that in law no one is to blame. It is particularly difficult for families of those who are killed, again through no fault of their own, by a person who was not in control of their actions. However, it remains unjust to punish someone for something they genuinely had no control over.

As was rightly highlighted by the shadow Solicitor General, the hon. Member for Torfaen (Nick Thomas-Symonds), the code for Crown prosecutors is clear about its dual test. The public interest, and the evidence and likelihood of securing a conviction, are the tests that Crown prosecutors will consider when assessing whether a defence is likely to undermine one or other of those factors. In that context, the hon. Member for Huddersfield mentioned transparency, as did the hon. Member for Torfaen. I think both will be aware that our system seeks to make the decision to prosecute as transparent as possible, including through communication with victims and those victims having the right to challenge and review the decision of the Crown Prosecution Service. In those cases, they will have information about what in those two tests caused the charge to not proceed.

Let me turn to the future, and the Law Commission’s report and proposals about this area. The lack of clarity on automatism to which the shadow Solicitor General referred, and the complexity of the outdated law on the connected defence of insanity, led the Law Commission to conduct a scoping study in 2012 and issue a discussion paper in the following year. The Law Commission did not, however, produce a full set of recommendations on automatism or complete its work with a final report. It rightly recognised that, in the context of its broader work on the law around insanity as a defence, this was an important but small part of a much broader piece. In taking that work forward, its focus was on the “unfitness to plead” aspect, and it is therefore yet to produce for consideration qualified legislative proposals on automatism.

Although the Law Commission’s comments and proposals in the 2013 discussion document would narrow the automatism defence slightly, as has been said, it would not remove it. The proposed reforms sought to simplify the law, replacing the common-law defence with a statutory one, and have one defence that, if made out, would lead to a not guilty verdict rather than the two possible verdicts previously mentioned. The Government considered the initial discussion paper’s proposals, but concluded that they would be a very limited reform to an already rarely used defence.

We have no current plans to bring forward legislative proposals. However, that does not mean that we have ruled out making changes to the law, including the wider and related law on insanity and fitness to plead; we keep that area under review. The shadow Solicitor General tempts me into saying that if the Ministry of Justice wishes to use any legislative time, that would be a good use of it. As a Department, we are productive in legislation; we have been, and I suspect we will continue to be.

To conclude, I congratulate the hon. Member for Huddersfield on drawing the attention of the House to this issue. It is, as I think he said, the first time in 11 years that it has been properly debated, and I am sure there will be Members who are hearing about this issue today for the first time. I thank all who have contributed to the discussion of this complex topic, and I hope I have set out the law as it stands and where the Government stand on it. Finally, the hon. Gentleman knows that we already have a meeting scheduled on a different topic. I enjoy my meetings with him, and if he wishes to add specific questions about this aspect of law to our next meeting, I am happy to consider them.