The new sentencing guidelines for health and safety offences published today are set to revolutionise punishment for health and safety offences.
The Sentencing Council’s stated intention is to increase the level of fines for serious offences, particularly for larger companies; whilst reserving prison sentences for very serious offences. In my view, the formulaic approach laid down in the new guidelines will greatly increase fines across the board and most dramatically for very large companies. More worryingly, many more directors, managers and junior employees will be handed custodial sentences due to a significantly lower threshold for imprisonment.
The new sentencing guidelines apply to health and safety offences committed by organisations and individuals, as well as to corporate manslaughter and food safety/hygiene offences. They introduce a structured nine step approach that the Court should follow, so as to calculate sentences. This involves plugging culpability and harm factors into a series of tables to reach recommended starting point fines, as well as ranges of fines above and below the starting points.
The new regime has been carefully calculated and will certainly improve consistency across England and Wales. It was published in draft for consultation in November 2014 and the final version published today is little changed from the draft, with the same figures and tables for fines as in the draft. These will apply to sentences handed down from 1 February 2016, which make the guidelines retrospective in the sense that they apply to offences that have already been committed and have not yet come up for sentencing.
My observations on the new guidelines are as follows:
The first step of the process is to allocate a culpability factor, ranging from ‘very high’ to ‘low’, which in turn will drive the allocation of a recommended fine. The ‘very high’ category is triggered if there has been a ‘deliberate breach of or flagrant disregard for the law’. Although that sounds very serious, with hindsight it may be all too easy for breaches to fall into this category.
It is not uncommon for employers to identify a safety problem and know they are non-compliant; this becomes a ‘deliberate breach’ if they don’t remedy it. If prosecuted, they may now all too easily find themselves at the sharpest end of the new regime.
The main health & safety offences are based on exposing people to risk. However, enforcement practice (as driven by public expectation) looks much more closely at outcome. Breaches causing serious injuries and death have until now been much more likely to be prosecuted than mere exposure to risk; and the sentences have also been much heavier.
The Sentencing Council has adopted a new rationale, dictating that punishment should much more closely fit the level of exposure to risk. So, exposing someone to a high risk of death should warrant almost the same sentence regardless of whether there is a fatality or merely an exposure to risk without any injury.
The consequence of this well intentioned change is to escalate offences that were previously dealt with moderately into the higher reaches of the new formulae. Imagine, a loose railing is discovered on a busy walkway that could lead to somebody falling from a height of 10 feet. The high risk of death or disability from someone falling can take this straight into the most serious harm category, even if nobody has fallen.
One of the steps is to designate the likelihood of harm as being ‘high’, ‘medium’ or ‘low’. That appears to be sensible. However, it may give false reassurance as to the fairness and objectivity of the new system. If 100 people walk along the unsafe walkway each day, risking a fall of 10 feet, is there a high, medium or low risk that someone will be killed or disabled?
The problem is that ‘high’, ‘medium’ and ‘low’ are prone to wildly different interpretations. Suppose the experts agree that there was a 1 in 10 risk that someone would have been killed or disabled by falling from that walkway throughout the period of the breach. Even with such a precise expert view, is that a high, medium or low likelihood? The court would struggle to answer that question without understanding the wider circumstances, looking at culpability and what actually happened. But those factors are intended to be excluded from this stage of the calculation to make the process rigorous and fair.
This issue will grab headlines in due course. The guidelines, in combination with the Court of Appeal’s July 2015 judgement against Thames Water Utilities, pave the way for the very large companies to be fined in excess of £100million for the most serious offences. That is clearly the Court of Appeal’s intention.
The culpability factor and harm category process is also used in order to produce a guideline starting point punishment for individuals. At the more serious end of the scale, the court is driven to impose a custodial sentence. So, where does the threshold for imprisonment now lie?
My interpretation of the new threshold is as follows: if a director or employee knows there is a breach of the law that has at least a medium likelihood of causing death or disability, then the court is directed as a starting point to impose a punishment of one year’s imprisonment, with a range of between 6 and 18 months depending on other relevant factors. Many will judge this to be fair if there has been a fatal accident. But alarmingly, imprisonment will be the conclusion even if there has been no accident at all – just exposure to risk.
This is a very significant reduction in the threshold for imprisonment for health and safety offences.
In summary, the sentencing guidelines adopt a range of well-intentioned and apparently rational changes. However, I am concerned the outcome will be very much more dramatic than intended, with much greater fines across the board and more individuals being imprisoned for offences that would not previously have been regarded as sufficiently serious to merit a custodial sentence.
It remains to be seen whether these changes will improve prevention and save lives, or drive hazardous industries abroad and bring public perceptions of over-zealous enforcement to new heights.
[…] Courts must also now look at the degree of culpability and the level of exposure to risk. This is a radical shift and it is no surprise that some experts in the health and safety industry agree that the Sentencing Council’s guide signifies “the most dramatic change in health and safety enforcement since 1974.” (*See SHP Online at https://www.shponline.co.uk/the-most-dramatic-change-in-health-and-safety-enforcement-since-1974). […]
You must be logged in to post a comment.
Simon, a very insightful article – thank you for sharing your views.
I’m tempted to think this fits into to the overall tendency for government and regulators “pulling back” from proactive to a do-if-you-dare culture, that if caught has significant punishment to deter others
Timing impeccable, I’m running a training course 04/11/15 for managers and directors, I’m using the article to support my slides and discussion. My guess is the delegates will be in utter disbelief, especially putting people at risk with zero accident or harm and being caught can lead to a custodial prison sentence.
I just hope that the medium to long term effect of these changes after several “big headlines” will NOT lead to army’s of ‘Safety-Bots’ crawling sites watching for the slightest error or lapse.
I think ‘dramatic’ is understating the effect, my thought is:- “eye-popping; jaw-dropping; heart-stopping; WTF.!!” is more likely.
I think HR teams will be busy reviewing company disciplinary procedures and I can anticipate the first overreaction being those new disciplinary procedures enacted against employees for minor safety infractions.
From a h&s practitioner’s perspective I think the new guidance is sensible and proportionate.
All too often sentences have been inconsistent and often paltry sentences have been given out for serious offences.
I particularly agree with focusing on the risk as opposed to the outcome of an incident.
The author of the article gives an example of a faulty railing at 10 foot where although no one has fallen the result is it becomes a high
risk offence. What the author has omitted is whether this condition was known and nothing was done to rectify it.
Furthermore if this railing was, for example, a tube scaffold which is installed to prevent workers from falling it is not unreasonable to
expect it to be fit for purpose. Even at 10 foot a serious injury, possibly life changing injury could occur.
However, increase the height to 10 metres and now the probability of a fatality increases significantly. So there is a good argument the greater the risk the more proportionate the controls to mitigate against that risk.
If someone should fall at either 10 foot or 10 metres the outcome is uncertain, good/bad fortune will dictate the actual outcome to some extent,
however the prevailing risk is apparent and constant. Another example from common law is someone firing a gun at someone but the bullet missing.
The culprit would undoubtedly be charged with attempted murder even though there was no fatality or even an injury.
In safety terms this person currently may not even get prosecuted, the enforcing authority is more likely to issue a Prohibition Notice for creating a dangerous condition,
which in some cases even this type of Notice has been ignored by the duty holder.
The author also states ‘ The problem is that ‘high’, ‘medium’ and ‘low’ are prone to wildly different interpretations.
While there is an element of subjectivity in risk assessment there are many risks which fall into the categories of low, meduim and high, which practitioners deal with on a daily basis.
It is not normally difficult to understand that certain types of hazards will result in a minor, serious, or fatal injury if a person comes into contact with them.
Even lay people will understand that if you come into contact with say, electricity, it is dangerous and the higher the voltage the more likely it is to kill.
Despite the belief that practitioners have a ‘crystal ball’ we do not know when, where or the outcome of an accident.
We do know, however, what constitutes to an unsafe or condition or act, and moreover, what is the probable outcome.
Thank you for your comments. Ray – you raise useful points. I chose 10 feet for my hypothetical example precisely because a fall from that height will most likely result in injuries less than fatal/disability, but nonetheless with a very real risk of fatality/disability. I am assuming that the failure in the railing is known about or ought reasonably to have been detected. Hence an offence has been committed.
You make a good point that safety practitioners are used to allocating risks as ‘low’, ‘medium’ or ‘high’ in risk assessments. That’s fine for a subjective opinion on some risk assessments but lawyers and judges are going to use this assessment in a purportedly objective manner to determine a punishment; that determination could lead to a difference of millions of pounds in fine level or, for an individual defendant, between a custodial sentence and a fine.
The guideline process requires a determination of pure likelihood of harm (on a low – medium – high scale), devoid from culpability which is supposed to be accommodated elsewhere in the calculation. Yet, if everyone in court accepts expert evidence that the risk was 1 in 10 that someone would be killed or disabled by falling from the walkway throughout the period of the breach, even though this likelihood has been so precisely determined, was that a ‘low’, ‘medium’ or ‘high’ likelihood? If we ask others to contribute to this debate, I expect we will get all three answers.
If I am right, then this aspect which lies at the heart of the new sentencing assessment is worryingly unreliable. I accept that sentencing on these cases will necessarily involve subjective views. What worries me is that judges may think they are following this part of the process in an objective way when that is inherently unlikely.
From an objective and scientific point of view, the guidelines should put numerical values to the likelihood levels in the tables. However, I can understand that this would have also caused confusion amongst lawyers and judges who may prefer words like ‘low’, ‘medium’ and ‘high’ without appreciating the inbuilt bias that engenders.
By moving away from an outcome based approach to sentencing that is easier to measure (e.g. there was a fractured ankle), these guidelines will take us into choppy waters that may only be good for lawyers arguing in very much lengthier sentencing hearings.
I would be most interested to get the views of a statistician on this and whether other readers regard a 1 in 10 risk as being ‘low’, ‘medium’ or ‘high’?
Hi Simon, thanks for responding. I do see the point you are making with regards to the precriptiveness of the process, which could lead judges to rely too much on an objective process when there is a great deal of subjectivity involved. This guidance must surely be used in conjuction with other evidence which I hope will provide the overall picture of the risk v outcome scenario.
I also meant to have mentioned how the ‘outcome’ can provide a negative effect for the accused. For example, in the case of Porter who was a headteacher prosecuted by the HSE because a pupil playing in the playground fell and injured his head on a set of steps. Unfortunately the young lad caught a MRSA infection and died as a result in hospital – tragic. However the HSE prosecuted Mr Porter, he was found guilty at first instance and with costs fined £20,000. Fortunately the CAP saw good sense and overturned the original judgment where the judge referred to the incident to an everyday risk. The reason for highlighting this is because I am convinced if the outcome had not been a fatality, then there is no way the HSE would have prosecuted the headteacher, nor would he have received such a large pecuniary sentence. There have been similar outcome based prosecutions – too may to list, but Blair v Condon springs to mind. With regards to your point of assessing a low, medium and high risk event, of course there will be differing views from so-called experts. That said, I don’t think they will vary wildly. A high risk may fall into a medium risk, but never a low risk category. With your example of a 10 foot fall a minor injury is probable, serious injury possible and fatal injury unlikely but not inconceivable – semantics maybe, but this is the language of h&s practitioners where very little is certain. As I commented previously, the outcome of an incident is often based on good/poor fortune. Why should someone be prosecuted and sentenced based on this?
Incidentally, your article has dropped off the main SHP page – so it’s unlikely many others will contribute, which is a shame.
Thanks, Ray. The Porter case was indeed a travesty until the Court of Appeal overturned it.
If anyone still has the energy to follow this debate, it takes us to a really important issue – the pros and cons of the existing predominantly outcome based approach to prosecution and sentencing as against the new predominantly risk based approach. There are indeed good reasons to justify a change of emphasis towards risk, which matches the HSWA offences. However, I believe the Sentencing Council has inadvertently lurched from one extreme to the other. The advantages of an outcome based approach are: it is easier to measure actual injuries as compared to risks; and society quite rightly demands tougher enforcement for worse outcomes.
Have you ever closed your eyes and nearly fallen asleep at the wheel? If so, by good fortune you and many other have woken up in time and avoided causing death by dangerous driving. But what about someone who does fall asleep and cause fatalities? The failing (driving whilst tired) is the same but the outcome is different. Most people agree the killer should be jailed and the rest of us lucky drivers should not. Society expects this outcome for good reasons, even though it may offend against a purist theory of risk and breach.
You didn’t tell me whether 1 in 10 is a low, medium or high likelihood? I ask because I don’t know the answer and am worried that I can precisely know the likelihood (1 in 10) and yet not know how to allocate it in the guidelines. Remember that the answer is not supposed to be dependent upon the context but is supposed to be a pure question in this pseudo-scientific approach.
Thank again Simon, as I implied, I do not believe the outcome of an incident can be measured objectively – there are too many variables. No one ever got injured falling from height, it’s when you hit the ground it hurts! From 10 feet landing on grass the resulting injury is very unlikely to be fatal, but a serious injury is not inconceivable. Landing on concrete or worse, a brick wall, palisade railings will undoubtedly result in a serious injury and possible fatal. Even if statistics are available (and I doubt they are) they would be of little worth becuase of the individual circumstances. Then there is the age and fitness of an indivdual, which would have a significant bearing on the severity of the injury. So, that is why I believe the risk element is such an important factor in assessing the culpability (if that is the correct term in this instance) as opposed to the outcome. The risk is constant in the case of an unsafe condition. The variables are factors of good/bad fortune. If you take another and perhaps better example – the risk of an electric shock. This is always a difficult outcome to predict for a number of reasons. First, it depends on whether the person is earthed if an injury is sustained, what part of the body the electricity passes e.g. the chest being the most danagerous, the voltage, impedence, prevailing weather conditions, AC or DC and so on. All of these factors will have a bearing on the outcome. However, if someone was to leave a live terminal in a unsafe condition they are exposing someone to the risk of an electric shock, fatal or otherwise.
Just a quick mention about the falling asleep scenario – I have done it! Another example on that theme is a train driver having a SPAD and causing a train crash (not done that – although I am an train ex-driver BTW). In the past the driver, assuming he was still alive…was prosecuted by the CPS if there was a fatality, common practice despite misgivings by HMRI. The last driver to be prosecuted was Peter Afford (Watford) partly because the drivers were not found gulity. No train driver deliberately goes past a danger signal, it’s inter alia human error. Thankfully we have a better understanding today of risk, human factors, accident causation, etc, where a laywer can explain to a jury a driver with x amount of experience does not suddenly become a reckless idiot! Indeed, the Cullen enquiry into the Ladbroke Grove train crash where both train drivers were killed, exonerated the driver who passed SN109? at danger. The reason? Mainly because it had been SPADed on numerous occassions and therefore is it was an existing unsafe condition. In fact he commented the occurrence of multiple SPADs at a signal “is a clear example of what is effectively an unsafe condition” and failure to solve the problem “represents an acceptance of that unsafe condition”. Res ipsa loquitur, so I rest my case. 🙂
Is not ironic that after pulling out The Management of Health and Safety at Work Approved Code of Practice (ACOP – L21) which was risk assessment prescriptive, the new sentencing guidelines focuses on the risk of harm as opposed to the outcome of harm!
Interesting times!!
Hello, I was on the webinar on Friday but have not received a copy of the slides. Please advise how I should get a copy of them or when they will be sent out to all who attended. Thank you very much, Juli Williams
Hi Simon,
As a H&S Officer of a Medium sized UK Ltd Company, I am interested to know your views on how this major change in implementation of tougher sentencing affects multi-national companies. Is the turnover of the parent company taken into consideration when calculating fines, or is it only calculated from the individual sub-company? If so, then this figure would rise into the Large Company category.
I would be grateful for your input.
Regards
Paul