Artificial agreements and unexpected outcomes

Written by Oliver Willmott

Introduction

Sentencing guidelines may not be tramlines – see R v Thornley [2011] EWCA Crim 153; [2011] 2 Cr.App.R.(S.) 62 and much other authority, before and since. Sentencing guidelines may not in every particular always receive universal acclaim amongst practitioners – see the varied reaction to the Health and Safety guideline. But there is an understandable consensus that they assist greatly in the process of advising as to likely sentence. Pre-guidelines there were only general principles to be distilled from authority and whatever help – usually not much – was to be had from reference to instances of the applications of those principles. Now the guidelines provide a structured process through which to approach the sentencing process; they define two dimensions of seriousness – harm and culpability; and in the case of the sentencing guideline on Health and Safety Offences they give harm two dimensions as well: the seriousness of the harm risked and the likelihood of the risk eventuating. Here is something for the lawyer advising to grapple with, something to show the defendant; here is a procedure to walk through, a toolkit for predicting sentence. Or that, at least, is the conventional wisdom.

I don’t seek here to refute conventional wisdom, but I do wish to reflect on it in examining one instance where, perhaps, a half-way accurate prediction of outcome would have required more than a knowledge of the facts, and a copy of the Guideline and a little logic. The case in question is Health and Safety Executive v ATE Truck & Trailer Sales Ltd [2018] 4 WLR 142; [2018] EWCA Crim 752.

The factual matrix

The defendant, ATE, dealt in trucks and trailers. Over a long period it had dealings with a scrap metal merchant. In the 1990s he had collected and sold on scrap generated when ATE refurbished or repaired vehicle trailers. In 1998 ATE started to convert curtain trailers to flat-bed trailers. It came to an arrangement with the contractor whereby he would remove the superstructure and sell it on for scrap. Where this process was undertaken varied over time, but by the time of the incident it was carried on in part of ATE’s premises, an eight-acre, open yard where it stored its trailers. The contractor paid £50 per structure in lieu of rent.

Tragically, on 21 February 2013 the contractor sustained head injuries in the process of dismantling a curtain-sided trailer which proved fatal. A health and safety inspector subsequently concluded that the contractor’s method of work was unsafe. ATE employed health and safety consultants and maintained a number of risk assessments, but it had no risk assessments in relation to the activity of dismantling curtain-sided trailers when conducted by ATE’s own employees and genuinely believed it had no responsibility for the contractor’s activities.

Proceedings at first instance

It seems that to begin with ATE was charged only with failing to discharge the responsibility imposed on employers by section 3(1) of the 1974 Act to ‘conduct his undertaking in such a way as to ensure, so far as is reasonably practicable, that persons not in his employment who may be affected thereby are not thereby exposed to risks to their health or safety’. But the prosecution subsequently indicated that a plea to a failure to provide a suitable and sufficient risk assessment for its own employees as required by regulation 3(1)(a) of the Management of Health and Safety at Work Regulations 1999 contrary to section 33(1)(c) of the Health and Safety at Work etc. Act 1974 would suffice. ATE entered such a plea and no evidence was offered in relation to the other charge and a not-guilty verdict entered.

The parties also agreed a basis of plea which not only dealt with the primary facts of the case but also how it should be characterised as a matter of the Guideline. Not every point, however, was agreed. The points of agreement and disagreement were as follows:

• ATE should have had a suitable and sufficient risk assessment in relation to the health and safety of its own employees when they undertook the dismantling of curtain-sided trailers.
• The method used by ATE employees was not at fault. The only failing, it was agreed, was the absence of a written risk assessment. The prosecutor said that this created the risk that it would not be followed. ATE said that the failure was a matter of form only.
• The offence was to be regarded ‘as having more than minimal or trivial connection’ to the incident in which the contractor died, but it was not a major cause of that incident. (see the Guideline at p. 5, point 2(ii)).
• The offence was one of ‘low culpability’.
• The harm risked fell into category A.
• ATE argued that the likelihood of the offence causing category A harm was low; the prosecution said it was medium.

The basis raises a natural question: how did the failure to have a written risk assessment for ATE’s employees in any way cause the accident involving the contractor? ATE explained the connection in this way:

‘Even though [the contractor] had the primary expertise in the activity he was undertaking and was extremely unlikely to pay any attention to the defendant’s own risk assessment, the defendant agrees it is possible he would have done so and adjusted his practices accordingly. Thus it is possible for the defendant to accept for the purposes of sentence that the failure to have a recorded risk assessment in relation to its own activity for its own employees can be regarded as having a connection … with [the contractor’s] accident.

[Emphasis added.]’

This is, to my mind, a somewhat extraordinary paragraph. ‘Unlikely’ becomes ‘possible’ becomes ‘admitted’. The test at p. 5, point 2(ii) is whether the offence actually was a significant cause of actual harm. What is needed is a cause, not a connection. It is difficult simultaneously to consider that it is unlikely that failing A was a significant cause of effect B and that it should be treated as proven fact that failing A was a significant cause of effect B. Doubtless ATE had sound, pragmatic reasons for proceeding in the way it did, but the two propositions appear to be destructive of each other, such that they cannot simultaneously be maintained.

The court at first instance expressed concerns as to the basis of plea in relation to causation, but was ultimately persuaded that he could sentence ATE ‘on a proper basis including causation’ of the contractor’s death. He believed it was wrong to ignore the concession even though he believed that there were ‘more appropriate and relevant charges that could have been laid against the company and which would have required the jury directly to consider the existence of any recognised legal duty of care towards [the contractor]’.

Critically, the Court did not accept the joint position of the parties as to how the case should be categorised as a matter of the Guideline:

• It did not agree that was actually being done by ATE was safe or that the want of a risk assessment was a matter of form. Without a risk assessment, no controls were in place. The work was inherently hazardous and such an assessment was plainly necessary. Many accidents were waiting to happen. The absence of an accident in the past could not be seen as confirmation of safety. The methods used by the contractor were clearly inadequate given the dangers in his work.
• Contrary to the agreement of the parties that culpability was low, it was actually high. ATE had failed to put in place measures that are recognised standards in the industry, because the industry would have expected both a risk assessment and a method assessment. Further, the breach had subsisted over a lengthy period – since 2002/3, it seemed.
• Harm was indeed at Level A, but the likelihood of harm was not medium as the prosecution maintained, or low as ATE suggested, but high. This work had been a substantial part of the contractor’s life, and he was at risk every time he carried it out. The offence created a high risk to all, and, by the concession as to causation, a high risk to the contractor in particular. Hence the case fell into harm category 1. Given that the basis of plea maintained that it was unlikely that the contractor would have had any regard to the missing risk assessment, had it existed, this was surely a bold conclusion.
• ATE had a turnover in 2016 of £17 million and accordingly was a medium-sized organisation, albeit at the bottom end. The starting point for a high-culpability offence in harm category harm was £950,000 with a category range of £600,000 to £2,500,000. The Court’s starting point was £625,000 (presumably this figure was arrived at by adjusting the figure of £950,000 to take into account the comparatively low turnover for a medium-sized organisation). This rose to £750,000 given that actual harm was caused to the contractor. Allowing for a plea of guilty and mitigation, that figure was reduced to produce a sentence of a fine of £475,000.

 The decision of the Court of Appeal

There was no doubt that the parties could not bind the court in relation to sentence: see R v Innospec Ltd [2010] Crim LR 665: ‘A private agreement between prosecution and defence will doubtless inform the court, but helpful though it may well be, cannot be determinative of sentence.’ Therefore, as a matter of principle, the judge was entitled not to concur in the agreement. But the question remained: was the decision to depart from it correct as a matter of fact.

The Court of Appeal noted the artificiality of the position as to concession but, like the court at first instance, declined to ignore it. Therefore page 5, point 2(ii) applied. But it followed that the court was bound to focus on the method of work of ATE’s own employees and the frequency of the work, not the method the contractor followed, or the frequency with which he followed it.

There was no, or no sufficient, justification for a finding that the case was one of high culpability. Consideration of the method followed by the contractor was outside the basis of plea and the relevant period for sentence was that specified on the indictment – a little over two years – not the period for which the contractor had been carrying on work. Save for the expectation of a risk assessment, there were no industry standards. Culpability was in truth low, as the parties had agreed.

The absence of a risk assessment in the context of dangerous work meant that the risk of Level A harm was not low. Nor did the absence of an accident suggest to the contrary: see R v Diamond Box Ltd [2017] EWCA Crim 1904 at paras 17–18. But the facts did not warrant classifying the case as one involving a high likelihood of category A harm. The Court of Appeal found that the judge at first instance had been influenced in his finding by the contractor’s method of work rather than ATE’s.

The Court therefore considered that the case fell into harm category 2 with low culpability. The Guideline gave a starting point of £40,000 with a range of £14,000 to £100,000. But owing to the fact that an actual death occurred – factor 2(ii) – it was right to move not only up a harm category, but to the top of that category. The next category had a starting point of £130,000 and a range of £75,000 to £300,000 and hence the figure reached was £300,000. This, though, fell to be reduced to allow for credit for a guilty plea. In the result, the sentence should be reduced from £475,000 to £200,000.

The Guideline and advice as to sentence

Against this background, I return to where I started – the way in which the Guideline permits practitioners to advise clients what fine they might receive in the event of conviction. How real is this supposed advantage of the Guideline?

In this case, the size of the enterprise was not in issue; the parties had agreed culpability at low and the harm category was (according to the prosecution) 2 and according to the defence, 3. That gave a starting point of £14,000 to £40,000 before credit for plea. If you allow for the court accepting the artificial agreement that the offence was causative of death and a move not only into the next category, but to the top of that category, those starting points become £100,000 to £300,000 before credit for plea. At that point advice might end. It could not possibly have predicted a situation whereby the offence was categorised as being one of high culpability and harm category 1, and thereby the sentence soared to £475,000. Even proceedings in the Court of Appeal where the starting point for sentence – absent the doubtful concession as to causation – fell as low as £40,000 before increased by a factor of seven-and-a-half to £300,000, prior to reduction for guilty plea.

Admittedly, the artificiality of the basis of plea was responsible for some of the volatility in sentence. But it was not the only factor. The category ranges are themselves wide, even once culpability and harm level are established and the width of the ranges varies depending on where an offence falls on the guideline. The Guideline may make sentences more easily predicted than was once the case, but levels of uncertainty are still considerable.