Halal meat must comply with certain requirements. They include that the animal is killed by having its throat cut so that it bleeds to death, and that the slaughterman must be a practising Muslim and recite a prayer. Halal certifying bodies differ on whether the animal must be conscious: most animals slaughtered in this way in the United Kingdom are stunned before being killed, by applying an electric current to the brain to induce an epileptic fit, but around 25% are not.

Euro Quality Lambs Limited (“EQL”) operates an abattoir that performs both stunned and non-stunned Halal slaughter. This appeal is about whether the way in which EQL performs non-stunned slaughter complies with the requirements in the Welfare of Animals at the Time of Killing (England) Regulations 2015.

After sheep arrive at EQL’s premises they stay in lairage, a broad term meaning the place where an animal rests on its way to slaughter. They are divided into pens according to the anticipated production order, for example grouped by size or age. EQL avoids keeping animals on its premises for any longer than necessary. If there is an unanticipated problem, such as sheep arriving too late on a Thursday to be slaughtered, they are placed in fields in a nearby farm over the weekend. But most sheep are slaughtered on the day after they arrive.

A sheep is a flock animal and will naturally follow the sheep in front of it. It is also a prey animal and (in this country) likely raised outdoors, so is wary of humans. As it was put by Fraser J in R. (The Association of Independent Meat Suppliers) v Secretary of State for Environment Food And Rural Affairs [2017] EWHC 1961 (Admin) (“AIMS)” at [22]:

So too indoors. If given a narrow run in a particular direction and prompted by a human from behind, sheep will follow one another in a line to their destination. In EQL’s premises, when a particular group of sheep are to be slaughtered, the correct number are let out of their pen and directed in that way to a crowd pen, as shown in this image.

It is a circular rotating pen, divided into segments. By opening and closing the entry gate to a segment of the crowd pen, and rotating it, a ring of sheep of the right density can be achieved that simply follows itself in a circle. On the right of the image, sheep can be seen entering the pen with a human operative behind them. The sheep in the foreground have been moved clockwise, halfway around the pen. The third and final segment exits to two sets of devices known as V-restrainers. In AIMS, Fraser J described a V-restrainer:

Sheep will naturally try to follow one another onto a V-restrainer, and the operative stood by the exit to the crowd pen assists and encourages those that do not. The operative can also slide a gate across to enforce single file and controls the turning of the pen. The utility of the crowd pen is that it enables sheep to be loaded onto the V-restrainers at a constant and predictable rate despite sheep arriving from lairage at irregular intervals. No sheep is brought into the crowd pen for any other purpose than to form part of the group to be killed in the next few minutes or hours. They would not, for example, rest there overnight. Sheep may have to wait in the crowd pen for a little while, depending on the speed of production, but as observed in AIMS at [38] they do so in a more conventional herd environment than if restrained in single file in a V-restrainer.

Each set of V-restrainers consists of three devices in series. The first V-restrainer entered by the sheep is the longest, and in stunned slaughter will hold around eight animals. In non-stunned slaughter only two sheep are loaded. The second and third V-restrainers follow immediately afterwards, and each holds only one sheep. Once two sheep have reached the end of the first V-restrainer, the belt pauses until the path ahead is clear. The sheep are then conveyed into the second and third V-restrainers, one in each. This means that one of the sheep passes through the second V-restrainer so that it can come to rest in the third. These two sheep, now stationary in separate V-restrainers, one in front of the other, then have their throats cut in turn. Each sheep’s head is manually lifted by a slaughterman who severs the carotid arteries and jugular veins with a sharpened knife. In the foreground of the image below is the third V-restrainer, behind it is the second V-restrainer, behind that is the first V-restrainer inclining downwards out of sight, and the grid to the left is where the slaughterman stands:

As recorded in AIMS at [10], following the incision a sheep must remain in position for at least 20 seconds while it loses consciousness and dies through exsanguination. Restraint of the sheep at the time the incision is made and for these 20 seconds is a requirement of the regulations. The slaughterman can then operate a control to convey the two dead sheep out of the front two V-restrainers so that they can be taken away for dressing, and to move the next two sheep out of the first V-restrainer to take their place.

As with the crowd pen, the efficiency of the system lies in providing a constant and controlled flow of sheep to the slaughtermen. The operative in the crowd pen can ensure that the first V-restrainer is loaded while the sheep in the second and third are dying. That efficiency is also shown in there being two sets of V-restrainers. The slaughterman alternates between them, spending the 20 second waiting time slaughtering a sheep on the other line.

It is only fair to record that EQL’s stated benefits for the process do not just include efficiency. EQL also puts it forward as more humane than the alternatives, for the reasons set out in AIMS at [12], [21]-[24] and in the evidence adduced in this appeal.

When the regulations were published, they were subject to a legal challenge by a trade association for abattoirs. In AIMS, the issue before the Administrative Court was the lawfulness of Paragraph 6(1)(a) of Schedule 3. For non-stunned slaughter, it provides as follows:

The claimant trade association relied upon evidence from one of its directions, Mr Rizwan Khalid, who is also a director of EQL. The practices considered by the Court were those of the very same abattoir with which this appeal is concerned, although they have altered in some respects since. Fraser J described the issues as follows:

He went on to rehearse the history of the legislation, including that it stood as the United Kingdom’s implementation of Council Regulation (EC) 1099/2009 of 24 September 2009 on the protection of animals at the time of killing (“the EU Regulation”), and the consultation and consideration of evidence conducted by DEFRA before the regulations were made. I take into account that full analysis without repeating it, only adding that the domestic regulations continue to apply as retained EU law.

The various grounds argued by AIMS were each rejected, the overarching conclusion being that Paragraph 6(1)(a) was rationally “aimed at ensuring more extensive protection of animals” within the meaning of Article 26 of the EU Regulation, DEFRA having established an underlying factual basis and reasoning such that its evaluative assessment could not be set aside on judicial review. There are other relevant parts of the judgment, to which I shall return.

Paragraph 6(1)(a) having emerged from the judicial review proceedings intact, attempts were made to enforce the legislation at EQL’s premises. The first step in enforcing this part of the regulations is to serve a Welfare Enforcement Notice, pursuant to regulation 38. It materially provides as follows:

A Welfare Enforcement Notice dated 27 November 2017 (“WEN 1”) was served on EQL by the Official Veterinarian assigned to its premises, Mr Mircea Pop. In a decision by Judge McKenna dated 5 July 2018, EQL’s appeal against WEN 1 was allowed. Judge McKenna interpreted regulation 38(2) as requiring that the opinion of the inspector be independently held. Having heard evidence from Mr Pop, she held that it was not, Mr Pop having simply served the WEN at the FSA’s behest. WEN 1 was cancelled.

The second WEN was dated 21 March 2019 (“WEN 2”). It was served by a different veterinarian, Mr Juan Chulian. The subsequent appeal against WEN 2 was unopposed by the FSA. Its reasons were not given, save that it intended to serve a third WEN. The appeal being unopposed, in a decision dated 14 June 2019 Judge McKenna allowed the appeal and cancelled the WEN.

Dated and served on 29 July 2019, this is the WEN with which this appeal is concerned. I shall refer to it simply as the “WEN”, save where it is necessary to distinguish it from WEN 1 and WEN 2. It was also served by Mr Chulian. The first relevant part of the notice reads that in the opinion of an authorised inspector, a contravention of Paragraph 6(1)(a) had been observed:

The next section reads:

This appeal was heard in-person at Birmingham Civil Justice Centre. It had been brought by EQL against the Food Standards Agency, and proceeded in that fashion up to and including the hearing. I have wondered since whether the correct respondent is actually Mr Chulian, but the parties have never raised the issue, nor has the FSA being the respondent been queried in other cases – see, for example, Food Standards Agency v Penny & others (European Union law, Other general regulatory appeals) [2022] UKUT 64 (AAC). It would be wrong to delay the resolution of this appeal further by requesting submissions on the point.

At the hearing, oral evidence was given by Mr Chulian and Dr Collin Willson on behalf of the respondent. Dr Willson is a veterinary surgeon and employed by the FSA as its ‘portfolio lead’ on Welfare at Slaughter. While Dr Willson was not called to give expert evidence, inevitably some of his factual evidence expresses the type of opinion discussed in Multiplex Constructions (UK) Ltd v Cleveland Bridge UK Ltd [2008] EWHC 2220 (TCC) at [670]-[672]. No objection was taken on behalf of EQL, but when assessing the weight to be afforded to this evidence I take account of the distinction. On behalf of EQL, evidence was given by Mr Khalid and an expert, Mr Peter Hewson. Mr Hewson is a veterinary surgeon with a distinguished curriculum vitae, and he retired from the FSA at the end of 2008, having spent his final year as its acting Veterinary Director. His expert report and its three addenda contain the appropriate confirmation of his overriding obligation to the Tribunal and understanding of his duties as an expert. The Tribunal also viewed videos that Mr Khalid had recorded of the slaughtering process in operation. I have not found it necessary to record any witness’s evidence in detail, but have taken it all into account.

The provisions entitling an inspector to serve a WEN are already set out above. By virtue of regulation 30(2), failure to comply with a WEN is a criminal offence. Regulation 39 entitles a person who is aggrieved by the service of a WEN to appeal against it. The powers of the Tribunal on appeal are set out in regulation 39(4):

This is worded in a similar way to the regulations considered in Health and Safety Inspector v Chevron [2018] UKSC 7. Lady Black held that while the opinion of an inspector was necessary to bring the notice into existence, on appeal the focus shifts away from that opinion to the notice itself. She continued:

I apply those principles. The Tribunal finds any relevant facts according to the standard of the balance of probabilities, and decides for itself whether the WEN should be cancelled, confirmed, or modified. Careful attention should also be paid to the reasons given for arriving at the decision under appeal, bearing in mind that Parliament has chosen (here) that an authorised inspector will be responsible for deciding whether to serve a WEN. The weight that ultimately attaches to the regulator’s reasons will depend on all these circumstances, taking into account their fullness and clarity, the nature of the issues, and the evidence: R. (Hope & Glory Public House Ltd) v City of Westminster Magistrates’ Court [2011] EWCA Civ 31 at [45]; Hesham Ali (Iraq) v Secretary of State for the Home Department [2016] UKSC 60 at [44]-[46].

The parties structured their submissions around EQL’s amended grounds of appeal, which can be summarised as follows:

I likewise set out my consideration according to the same headings.

There is no real factual dispute under this head, rather disagreement on how the statutory language should be applied to the way in which lambs are slaughtered at EQL’s premises.

Despite EQL’s dissatisfaction at the service of a third WEN, for the purposes of Ground 1 the respondent has, I consider, maintained a consistent position. Paragraph 6(1)(a) prohibits the placement of an animal in restraining equipment unless the person who is to carry out the killing is ready to make the incision immediately after it is placed in the equipment. At the point a sheep enters the first V-restrainer it is restrained, but the slaughterman is not ready to make the incision immediately afterwards. Even if only two sheep are placed in the first V-restrainer at once, the slaughterman is likely busy slaughtering the previous two sheep or waiting for the compulsory 20 second period. The sheep must wait its turn, in restraint, before the slaughterman is ready to make the incision. It is the same case as was made by DEFRA in AIMS, and discussed by Fraser J at [11]. The WEN can also be fairly read as complaining that sheep pass through more than one restrainer before being killed.

EQL’s case focuses on the parts of Paragraph 6(1)(a) to which I have supplied emphasis below:

The requirement, argues EQL, therefore only applies once it is “engaged in the killing of a sheep”. The phrase denotes a process rather than a one-off act, and can be contrasted with Paragraph 5 that refers to any “person who kills a sheep…”. The most natural meaning of the process of killing is that it begins upon the slaughterman’s knife severing the carotid artery. It is then that the process of killing begins, and EQL points to expert evidence that the sheep loses consciousness some 7 to 10 seconds later, and that death occurs at around 14 seconds. So, EQL protests, it is no more open to the respondent to describe loading the sheep into the first V-restrainer as part of the killing process than when it is led into the crowd pen, or the lairage, or (I suppose) if it were physically restrained at the farm on failing to cooperate with being loaded into the livestock trailer, or any of the other various steps that eventually result in the sheep’s demise. No sheep is killed in the first V-restrainer, which simply transports a sheep from one part of the premises to another. The restraining device for the purposes of killing the sheep, argues EQL, is the V-restrainer in which it is actually killed.

At Regulation 3(2), the domestic Regulations explicitly adopt the definitions in the EU Regulation. So, EQL points to the following definitions at Article 2:

Next, EQL points to the requirement at Article 15(2):

EQL argues that this provides a complete picture of the process: stunning, if performed; restraint, mandatory if the animal is not stunned; and killing. It would be absurd, say EQL, to suggest that ‘stunning’ as so defined takes place anywhere other than the final V-restrainer in which the live sheep comes to rest: that is where the electrodes are placed on its head. Likewise, killing can only be said to occur in the final V-restrainer: that is where the ‘intentionally induced process which causes the death’ of the sheep is started, being the severing of its carotid artery by the slaughterman’s knife. On the respondent’s interpretation, the process of killing the animal begins before it is stunned. EQL argues that cannot be right.

In support of EQL’s arguments, Mr Thomas put to the respondent’s witnesses that the ‘process which causes the death of an animal’ begins with the application of the knife to its neck. Mr Chulian replied that he saw it as an individual step in the whole process, which began when the sheep was placed on the first, inclined, V-restrainer, and that otherwise the phrase ‘immediately after’ in Paragraph 6(1)(a) would serve no purpose. Dr Willson agreed that, in literal terms, killing the animal begins with making the incision and ends 15 to 20 seconds later. Whether the ‘process’ of killing could begin sooner, and at what stage, he described as a philosophical question.

Article 9(3) of the EU Regulation requires that:

It is difficult see how this can be read as doing anything other than regulating what happens in the moments before incision, not from incision. On the domestic regulations, Fraser J held as follows:

Arguably at least, this is conclusive. Fraser J holds that the wording of Paragraph 6(1)(a) means the same as “each sheep has to be killed immediately it is restrained”. At [53], he further held that it was lawful for Paragraph 6(1)(a) to impose stricter requirements than the EU Regulation. Even if I am not formally bound by either conclusion, I agree with them. On any rational view, sheep in EQL’s process are not killed immediately after being restrained. They are restrained from the moment they enter the first V-restrainer, and must wait in it until the sheep in front of them have been killed. The longest a sheep did so in Mr Khalid’s video was 52 seconds. It was not argued that this met the requirement of being “immediate” in Paragraph 6(1)(a), nor could it: see AIMS at [39].

I nonetheless address the issue as it was argued. EQL’s suggested interpretation of Paragraph 6(1)(a) has not remained entirely consistent. In the amended grounds, EQL argues that the provision “applies to handling during killing”, and that “the killing process begins when the knife is put to the animal’s neck, which is then cut to sever the carotid arteries that supply blood to the brain, and ends when the animal is dead from blood loss”. EQL argued categorically that “the legislative clause thus does not apply to any process, or part of a process, that occurs before then (such as moving the animals from the motorised crown pen through [the first V-restrainer]”. This was the case put to the respondent’s witnesses in cross-examination.

In Mr Thomas’s skeleton argument this argument is repeated, but with additional focus on the different V-restrainers. The only device in which the animal is restrained for the purposes of killing, he argues at paragraph 51, is the one in which it comes to rest: the animal is neither killed nor restrained for the purposes of killing in the first device.

Challenging EQL’s case that the requirements only apply from the point of incision, Mr Ostrowski asked Mr Khalid how a slaughterman could ever not be ready to make the incision immediately, given that he was already making it. Mr Khalid’s response was that the ‘process of killing’ actually begins when the sheep arrives in the V-restrainer in front of the slaughterman. This was echoed in Mr Thomas’s closing submissions, where he described EQL’s case as being that the process of killing begins when the animal ‘comes to rest’, at which point the slaughterman is ready to immediately make the incision. I cannot accept that this is consistent with EQL’s case in the amended grounds. Either the requirements at Paragraph 6(1)(a) apply from the sheep coming to rest or they apply from the point of incision. It cannot be both.

None of the above should be taken as criticising Mr Thomas, who skilfully and persuasively argued EQL’s case, nor Mr Khalid, who I am satisfied gave entirely honest evidence. The shift in focus simply reveals that EQL’s primary interpretation cannot withstand analysis. Incision is the point from which the sheep must be restrained, pursuant to Article 15(2) of the EU Regulation. If that is also the point at which Paragraph 6(1)(a) begins to apply, then its obligation not to restrain could never arise. The provision would be rendered redundant.

The Tribunal can still address EQL’s alternative case, being whether the requirements of Paragraph 6(1)(a) commence on the sheep coming to rest in its final V-restrainer. I accept, from the videos of the process, that the slaughterman is ready to immediately make the incision when the sheep finally comes to rest, but reject EQL’s argument that this is the first time the sheep is restrained for the purpose of being killed. The argument depends on the three V-restrainers being seen as separate incidents of restraint. In AIMS, Fraser J appears to have contemplated that there would be only one V-restrainer, and I infer that placing three devices in series is a later development. In addressing the lawfulness of restraining multiple sheep in a single V-restrainer, he held as follows:

It might be wondered, after reading those extracts, how EQL can argue that its system of having four sheep restrained at a time is lawful. Again, its case depends on the first V-restrainer in the series falling outside the scope of Paragraph 6(1)(a). I reject this. The prohibition is against placing a sheep in “restraining equipment”, not against placing a sheep in an individual device. It is entirely consistent with the statutory language for three v-restrainers in continuous series to together form “restraining equipment”, into which the animal cannot be put until the slaughterman is ready to make the incision immediately, and I make such a finding here.

The above point can be further illustrated, were it necessary, by putting oneself in the place of the sheep. It finds itself subjected to only one period of restraint, undergoing an uninterrupted experience that meets the definition of Article 2(g) of the EU Regulation on its entire journey through the three V-restrainers, just as if it were placed in a stationary cradle then made to wait up to 52 seconds before the slaughterman is ready to make the incision. It is unreal to suggest that placing the sheep in the first V-restrainer is not for the purposes of killing, when it commences the same action of restraint in which the sheep is ultimately killed. Any enhancement to the sheep’s welfare from being behind another sheep does not undermine this conclusion, and in any event was considered in AIMS when deciding that the regulations were lawful as drafted.

I therefore dismiss EQL’s appeal on this ground without needing to address the respondent’s further arguments in detail. Nonetheless, it is right to set out a brief assessment:

Regulation 38(2) provides that:

So, a WEN may only be served by “an inspector”, being a person who has been appointed by (in this case) the FSA pursuant to regulation 34. The FSA cannot serve a WEN itself. While it appoints inspectors, it does not have a direct employment relationship with them. Instead, inspectors are employed by a private company called Eville & Jones, operating on the FSA’s behalf. Judge McKenna held that the opinion required by regulation 38 must be an “independent” opinion, but her judgment must be seen in the context of the unsatisfactory evidence before her, leading to findings that Mr Pop had simply served the notice at the behest of others, without forming his own view. I reject that she intended to place any gloss on the provision, which should be approached according to normal public law principles.

In the amended grounds of appeal, this second ground makes multiple assertions of non-compliance with the statutory scheme. These were narrowed significantly in Mr Thomas’s closing submissions, and can now be approached as an argument that Mr Chulian’s opinion was insufficiently independent to meet the requirement of regulation 38(2). This is said to arise in two linked ways. First, EQL argues that an impermissibly ‘top-down’ approach was taken to approaching compliance. If DEFRA were of the view that a particular practice contravened the legislation, this guidance was cascaded down to inspectors such as to strip them of the independence intended by the legislation. Second, this independence is further undermined if the FSA or Eville & Jones (and perhaps DEFRA) can simply send any inspector of their choice to serve a WEN. They argue that a WEN must instead be served by the Official Veterinarian (“OV”) actually assigned to the premises, rather than an inspector who has been hand-picked for that purpose.

At the time he served the WEN, Mr Chulian was a Deputy Area Veterinary Manager for Eville & Jones and appointed as an inspector under regulation 34. Approved premises, such as those operated by EQL, each have their own assigned OV. This enables a direct relationship between the OV and the managers of the premises; issues of concern can be discussed and tracked, as well as their improvement. Mr Chulian represented the layer of management above OVs.

The first WEN had been served by the OV assigned to EQL’s premises, Mr Pop. Mr Chulian explained that numerous WENs had been served on different premises following the judgment in AIMS, and that this had been done by their OVs. After the first WEN served on EQL had been successfully appealed, it was decided that the legal issues warranted involvement of a manager. Mr Chulian had previously worked as the OV at EQL’s premises himself, so was considered the most suitable person.

Despite the passage of time since the events in question, a subsequent change of employment, and comprehensive cross-examination by Mr Thomas, Mr Chulian’s evidence was careful, well-compiled and consistent. He struck me as a thoughtful witness, willing to analyse his and others’ actions candidly and critically. Mr Thomas conceded that Mr Chulian’s evidence was credible, and that he had genuinely held an opinion that there was a contravention.

Prior to serving the second WEN, Mr Chulian had seen EQL’s process in operation on around 13 occasions, as part of his OV and management duties and as part of his consideration of the present issue. He accepted that he had not held any concerns about EQL’s process on most or all of those occasions. His view of the requirements had changed following the publication of the regulations, the case of AIMS, different iterations of DEFRA and FSA guidance, and conversations and training in his workplace. I pause to observe that significant evidence was provided of these matters, particular in the evidence of Dr Collin Willson, who had a senior role in the FSA’s handling of the issue and wrote much of its guidance. It does not all need repeating in these reasons, but it is fair to say that the perceived requirements of the legislation, and whether businesses such as EQL complied with them, was a rapidly changing picture.

The challenge put to Mr Chulian in cross-examination was that he had simply followed guidance received from above instead of reaching his own independent view. He denied this, but candidly accepted that he had regard to guidance and legislation as setting out the basis for his own views on animal welfare. He had taken part in meetings with others from Eville & Jones, and the FSA, on the issue of whether EQL complied with Paragraph 6(1)(a). He had not, as EQL initially suggested, been compelled (in the sense of receiving a lawful instruction from an employer) to visit the premises and serve a WEN. He had been fully in agreement with both the decision that he be the inspector, and had formed his own view upon visiting the premises that EQL was in contravention. While responsibility for serving notices was that of an individual inspector, he valued working as a team when addressing particular issues.

The credibility of Mr Chulian’s evidence was accepted by Mr Thomas. That evidence included the following:

I have no hesitation in finding that Mr Chulian approached his task in that way.

Consequently, this ground concerns only whether the opinion formed by Mr Chulian satisfies regulation 38(2). That opinion concerns whether a person has contravened or is contravening the regulations or the EU Regulation. Some of the arguments put forward by EQL tended to conflate that issue with whether a particular practice gives rise to animal welfare concerns more generally. That is not a necessary ingredient. An inspector is concerned with assessing what is happening, and then forming an opinion on whether it contravenes the legislation. For example, Paragraph 6(2)(a) prohibits moving an animal for at least 20 seconds after the cut is made. If an inspector was of the personal opinion that only 15 seconds is necessary to safeguard the animal’s welfare, that would be no bar to him issuing a WEN in response to a practice of only allowing 18 seconds. What matters is the legislation.

The relevance of the inspector’s opinion on animal welfare would only be potentially relevant to (a) interpretation of legislation, where ambiguous, (b) where specifically incorporated by a particular provision or (c) the exercise of discretion as to whether the WEN should be issued. For the reasons already given in relation to ground one, I do not consider that issues of animal welfare can make a material contribution to interpreting Paragraph 6(1)(a) or deciding whether it has been contravened. Animal welfare considerations might, in some cases, be relevant to discretion; as observed during evidence, not all breaches justify taking formal enforcement action. That is, however, a separate issue to whether the opinion required by regulation 38(2) has been established.

The best fit for EQL’s arguments, in my assessment, is the concept of bias or predetermination: that Mr Chulian has been told what to think and / or approaches the decision with a closed mind. The test was expressed as follows in Persimmon Homes Teesside Ltd v R. (Lewis) [2008] EWCA Civ 746:

That authority (referred to as Lewis v Redcar) was applied in R. (Fraser) v National Institute for Health and Clinical Excellence [2009] EWHC 452 (Admin), the facts of which bear some similarity to those in this appeal. The Claimant had argued that some of the members of a committee had been nominated according to an impermissible preference for a particular form of treatment, and had then carried forward that predetermination into the decision under challenge. Like Mr Chulian, they were accused of having been both chosen for their particular stance, and of having predetermined the issue. After summarising the relevant authorities at [49], Simon J held that:

In the present appeal there is no direct evidence, or evidence from which an inference could fairly be drawn, that Mr Chulian had a mind that was closed to the issue. On the contrary, I find that he formed his own independent view. Nor, having regard to the overall context of the scheme by which inspectors are appointed and carry out their work, is bias or predisposition demonstrated by him being the particular inspector who served the WEN. Simply gaining his understanding of the law from guidance and training provided by the FSA or DEFRA cannot show bias or predisposition. The law is replete with statutory officeholders who must independently apply laws on which they have received guidance and training, from police officers to judges to health & safety inspectors. Independence does not mean closing one’s ears, it instead demands the opposite. If, hypothetically, an inspector were told by their employer “you must serve this notice or you will be dismissed”, then that might, on the particular facts, undermine his opinion. That is not the case in this appeal. Having heard the FSA and DEFRA’s view of the law and EQL’s rebuttal, Mr Chulian agreed with the former.

Nothing in the statutory scheme restricts the ability to serve a notice to the premises’ own OV. In any event, EQL was unable to identify any formal way in which an OV is ‘assigned’ to particular premises, and accepted during argument that this might change from day to day to address sickness and annual leave. That is a matter of employee management for Eville & Jones to determine, subject only to any relevant provision of its agreement with the FSA. An OV is not assigned to premises by way of a formal legal decision, like a First-tier Tribunal Judge to a Chamber, it is simply a working arrangement. There is certainly no basis for elevating that arrangement to a statutory restriction on who may serve a WEN, and in many regulatory fields it will be appropriate for an inspector in a particular position, or with specialist knowledge, to serve a particular notice. Nor is there any basis upon which to consider that Mr Chulian unlawfully delegated his statutory responsibility, or served the notice acting as the agent of Eville & Jones or the FSA.

Regulation 38(3) mandates particular information to be included in a WEN. The parties agree that if the WEN under appeal fails to comply, the appeal will succeed.

EQL argues that the WEN fails to comply with regulation 38(3), in that it fails to:

EQL’s grounds argue that a WEN must tell the recipient “precisely what is alleged to be the breach or non compliance and also to know precisely what it is being asked to do”, and that this should be “without the need to refer to other documentation or information.” Likewise, Mr Thomas argued that the recipient must be informed “simply by reading the WEN, what law it is alleged to be contravening, or has contravened, why the inspector has formed that opinion and what the inspector is telling the [recipient] to do.”

Mr Thomas cited Miller-Mead v Minister of Housing and Local Government [1963] 2 QB 196, CA, a case concerning an enforcement notice issued under the provisions of the then-applicable Town and Country Planning Act 1947. On the content of such a notice, Upjohn LJ held as follows:

Mr Thomas also referred me to the application of that authority by HHJ Waksman KC in Oates v Secretary of State for Communities and Local Government [2017] EWHC 2716 (Admin). In considering an argument that the enforcement notice in that case was a nullity by reason of deficient wording, the Judge held as follows:

Mr Ostrowski cautions against over-reliance on the judgment in Miller-Mead, given subsequent changes to the statutory scheme for planning enforcement, citing several more modern authorities in support. I agree, and note that in Oates the Judge preferred the phrase “could not tell with reasonable certainty what steps he had to take”.

I question whether the clear language of regulation 38(3) requires elaboration at all, and whether relevant guidance can really be drawn from authorities concerning enforcement notices issued under planning legislation. Nonetheless, asking whether the recipient can “tell with reasonable certainty what steps he had to take” is a useful way of asking oneself whether those steps have been specified, as required by the present legislation. I reject any argument, and EQL’s grounds of appeal could be interpreted as going that far, that the nature of the contravention must be entirely and precisely definable from the four corners of the notice itself, perhaps even to an ignorant third party. That would ignore both the ‘reasonably’ part of the test and that it is the recipient who must be able to tell what steps to take. If the actual recipient of the notice can tell with reasonable certainty what matters are said to contravene the legislation, and what to do to remedy that contravention, that is sufficient. Finally, I reject that the regulation requires any particular reasoning be given in support of the inspector’s opinion, let alone how that opinion was formed. The WEN is not a pleading, and need not specify any other information than is explicitly required by regulation 38(3).

On the WEN itself, Mr Thomas attempted to illustrate its non-compliance by dissecting the wording. The contravention is specified as follows:

On the basis, he argued, that the first two sentences must be introductory, only the final sentence can be taken as specifying the contravention, and it fails to do so with the requisite precision. I disagree. The various matters put forward by EQL as establishing a lack of precision are those it has already fielded in support of ground 1: the killing process only starts on incision, the first V-restrainer is not used for the purposes of killing and is not the same restraint in which the animal is killed, and so on. If the WEN makes no sense on EQL’s interpretation of Paragraph 6(1)(a), that says rather more about EQL’s interpretation than it does about the notice. The notice could hardly be clearer that, using the language of these reasons, the inspector is of the opinion that placing the animal in the first V-restrainer when the slaughterman is not yet ready to immediately make the incision is unlawful. EQL does not pretend to be ignorant as to the distinction drawn in the notice between the feeder restrainer and the restrainer in which an animal is to be killed, which is clear on both its face and in context.

On the steps required, the notice specifies the following:

I reject EQL’s contention that this fails to meet the requirement at regulation 38(3). It had been told what it was doing was wrong, and the notice told it to stop. That is enough. As argued by Mr Ostrowski, there is no duty to provide EQL with advice on how to implement a system that does comply. An injunction not to do something in a particular way does not require instruction on how to do it in another way, unless failing to do so would make compliance impossible. The remainder of EQL’s arguments on this topic, relating to animal welfare and the practicability of a non-restrained ‘race’ to the restrainer in which the sheep is killed, do not assist it. The former amounts to further disagreement with the merit of the legislation, already addressed in AIMS; and on the latter, the evidence establishes that by the time the WEN was served, EQL were fully aware of how other premises were conducting non-stunned slaughter that met the requirements of the regulations.

The WEN is compliant with regulation 38.

Some of EQL’s arguments on this ground, such as the lack of an independent opinion, have already been addressed. This leaves an argument that the WEN should be cancelled due to it being the third attempt at enforcement, the two previous WENs having been cancelled by the Tribunal on appeal. Mr Thomas relied on the principle of Henderson abuse. In Johnson v. Gore Wood & Co [2000] UKHL 65, Lord Bingham approved the description of the rule given in Barrow v. Bankside Agency Ltd [1996] 1 WLR 257:

I was not referred to any authority for the proposition that a party may rely upon Henderson abuse in an administrative tribunal. There is some discussion of the whether the doctrine of issue estoppel, a similar concept, applies in judicial review proceedings by Kerr J, in R. (Fire Brigades Union) v South Yorkshire Fire And Rescue Authority [2018] EWHC 1229 (Admin) at [85]-[88]. In this appeal, I consider that a Tribunal is entitled to take into account its own previous decisions in the way discussed in Secretary of State for the Home Department v BK (Afghanistan) [2019] EWCA Civ 1358, particularly at [39]. If an appeal against a WEN were to be allowed on the basis that the impugned procedure did not contravene the substantive requirements of the legislation, then an appeal against a subsequent identical WEN could be allowed on the basis that the issue had already been resolved.

In this appeal, whatever the precise juridical basis for addressing the issue, I reject that the previous enforcement activity justifies the Tribunal cancelling the present WEN. First, neither of the previous appeals resolved whether EQL is in contravention of the regulations. That is a matter of importance, and there is a public interest in the real dispute between the parties being decided. Second, as well put by Mr Ostrowski, if this WEN is quashed on abuse grounds then EQL will be left immune from a regulatory requirement that will continue to apply to others. While Henderson abuse might stand as grounds for denying a private remedy due to the way in which it had been pursued, I have not been referred to any authority in which it stood as justification for permitting indefinite future illegality.

EQL’s grounds of appeal each fail. The appeal is dismissed and the notice confirmed without modification.