England and Wales High Court (Administrative Court) Decisions

IN THE HIGH COURT OF JUSTICE
KING'S BENCH DIVISION
ADMINISTRATIVE COURT
SITTING IN LEEDS

		10th November 2023

FORDHAM J:

Part 1: Introduction

1. This is a case about the costs of an abandoned crown court trial and the limits of the supervisory jurisdiction of the Administrative Court. The case raises two questions for me to decide. One is whether the costs order made by a crown court judge involved a public law error, as is claimed. The other, legally prior, question is whether that claim falls within the Administrative Court's supervisory jurisdiction.
2. The Crown Court is the defendant to these judicial review proceedings. However, as is conventional in these cases, it has not participated. As Ms Otter from the Crown Court put it in an email response to the claim: "Whether the challenge to the decision of the judge in the course of criminal proceedings has merit is a matter for determination by the High Court". The active parties have been Exolum and the HSE who were the sole parties – defence and prosecution – in the Crown Court.

Regulation 3

3. The relevant power, to make the crown court costs order which is challenged in this claim for judicial review, is found within Regulation 3(1)(b) of the Costs in Criminal Cases (General) Regulations 1986. Regulation 3(1)(b) provides:

Subject to the provisions of this regulation, where at any time during criminal proceedings – … (b) the Crown Court … is satisfied that costs have been incurred in respect of the proceedings by one of the parties as a result of an unnecessary or improper act or omission by, or on behalf of, another party to the proceedings, the court may, after hearing the parties, order that all or part of the costs so incurred by that party shall be paid to him by the other party.

Regulation 3 was made pursuant to s.19(1) of the Prosecution of Offences Act 1985, from which its language derives. The relevant procedure is set out in Criminal Procedure Rules 45.8. There is a recommended 3-staged approach set out in Criminal Practice Direction Part 45 §4.1.1.

Components of Regulation 3

4. The costs ordering function in Regulation 3 has the following key Components. (1) First, there is a Default Precondition. This is reflected in the language "is satisfied" and "an unnecessary or improper act or omission by, or on behalf of, another party to the proceedings". (2) Secondly, there is a Causation Precondition. This is reflected in the language "is satisfied that costs have been incurred in respect of the proceedings by one of the parties as a result of…" (3) Thirdly, there is a discretion. It arises where the Default and Causation Preconditions are satisfied. It is reflected in the language "the court may order". (4) Fourthly, there is a Procedural Fairness Guarantee. It is reflected in the language "after hearing the parties". (5) Fifthly, there is a Specification Duty. This duty is found in Regulation 3(3), which provides that a Regulation 3(1) costs order "shall specify the amount of costs to be paid in pursuance of the order".
5. These Components of Regulation 3 feature in some of the cases to which I will return. In CCE (Leicester) there was consideration of the Default Precondition, the Causation Precondition, and the Specification Duty (§22 below). DPP (Sheffield) concerned the Default Precondition (§32 below). DPP (Aylesbury) concerned the Default Precondition and the Specification Duty (§36 below). This case is about the Causation Precondition.

Part 2: Jurisdiction

6. HSE agrees with Exolum that – were I to conclude that the Judge's approach to the Causation Precondition involved a material public law error (an "error of law", as Mr Puzey put it) – this Court would have jurisdiction to grant judicial review and make a quashing order. That common position was based on the Harrow (Maidstone) line of authorities, to which I will turn. But the Court needs to satisfy itself as to a question of its jurisdiction. The parties cannot confer jurisdiction by agreement. Permission for judicial review has been granted, but there has been no determination of a preliminary issue as to jurisdiction. So, I have to grapple with this legally prior question. It was helpfully addressed in writing and orally by Mr Molyneaux and Mr Puzey.

Public Law Error

7. Judicial review is a common law supervisory jurisdiction which has evolved to secure the constitutional imperative of accountability to the rule of law. It is a jurisdiction over public authorities, including 'inferior courts' of 'limited jurisdiction'. It operates to enforce a suite of well-established grounds for intervention. That is the product of the painstaking incremental development of the common law. These grounds for intervention do not extend to a substitutionary review on the merits of decisions on matters of evaluative judgment entrusted to the public authority as primary decision-maker. The grounds can still fit within a broad classification into unlawfulness (substantive errors for substitutionary review), unreasonableness (substantive errors for non-substitutionary review) and procedural unfairness (errors which are procedural). A public law error will only vitiate a decision if it is a material error. There are unifying descriptions used to encompass the entire suite of judicial review grounds. They include the phrases "public law error" (see eg. R (Lumba) v SSHD [2011] UKSC 12 [2012] 1 AC 245 at §70) and "public law wrong" (see eg. R (Privacy International) v Investigatory Powers Tribunal [2019] UKSC 22 [2020] AC 491 at §173). In the same way, all grounds for judicial review fall within the statutory phrases "point of law" (see eg. James v Hertsmere Borough Council [2020] EWCA Civ 489 at §31) and "wrong in law" (see eg. Taylor v Solihull MBC [2020] EWHC 412 (Admin) at §40). In this judgment I will use the phrase "public law error".

Recognised Species of Public Law Error

8. In the light of the arguments and the analysis which are to follow, it is helpful at the outset to identify the following as included within the established species of public law error: (i) failure to ask the legally required question; (ii) error as to a precedent fact; (iii) an unreasonable decision; (iv) a conclusion unsupported by any evidence; (v) misdirection in law; (vi) exercising a statutory power other than for the purpose for which it has been conferred; (vii) failure to make a required finding; (viii) insufficiency of enquiry; (ix) failure to discharge a duty; and (x) a decision made in reliance on a legally irrelevant consideration. All of these will be encountered later, in the discussion of the Harrow (Maidstone) line of authority.
9. Exolum contends that the Judge made a public law error in his approach to a statutory precondition to the crown court's "jurisdiction" to make a Regulation 3 costs order. That statutory precondition is the Causation Precondition. As Exolum's skeleton argument emphasises, the public law errors claimed all relate to the "jurisdictional threshold" of the Causation Precondition. As I saw it, Exolum's claim is variously put as: an error of law as to whether the precondition was satisfied; a failure to ask a legally required question; the failure to make a required finding; an unreasonable decision; a conclusion unsupported by any evidence; the giving of legally inadequate reasons; and a breach of a statutory duty (to act compatibly with Article 1 Protocol 1 rights under the Human Rights Act 1998). All of these are established species of public law error. There is no difficulty in the case being put in different ways. Grounds for judicial review frequently overlap, as in this case.

Modified Review

10. Judicial review is not necessarily an 'all-or-nothing' jurisdiction. There can be situations where some grounds for judicial review are unavailable, or where a particular test has to be met. The classic example is R (Cart) v Upper Tribunal [2011] UKSC 28 [2012] 1 AC 663, where judicial review was held to be available in relation to the Upper Tribunal's refusal of permission to appeal, notwithstanding the statutory overlay and the UT's designation as a superior court of record. But it was judicial review only on a restricted basis requiring not only a public law error but also (a) an important point of principle of practice or (b) some other compelling reason to hear the case. An example from pre-Cart case-law is R (Strickson) v Preston County Court [2007] EWCA Civ 1132 at §27 (identifying the need for a "pre-Anisminic" jurisdictional error or "grave procedural irregularity"), §32 (asking whether "the judicial process itself has been frustrated or corrupted") and §34 (leaving open whether judicial review would lie for error of law on the face of the record).

Statutory Overlay and the Rule of Law

11. In the context of the crown court, there is a statutory overlay to the judicial review jurisdiction. The law applicable when Parliament has enacted a statutory overlay – including where there is the prospect of the statutory exclusion or curtailment of judicial review – is reflected in Cart (including in the Divisional Court [2009] EWHC 3052 (Admin) [2011] QB 120: Cart (DC)) and in Privacy International. But those were not cases concerned with the crown court. Parliament's statutory overlay can undoubtedly influence the scope and shape of judicial review. That is clearly seen in the modified review outcome in Cart. The final arbiters of whether and how that operates are the Courts. And there is a constitutional touchstone: the need to secure the level of scrutiny required by the rule of law: see Cart (SC) at §51 (Lady Hale), §89 (Lord Phillips), §122 (Lord Dyson); Privacy International at §§131-132. And it is the Courts who determine what the rule of law requires: Privacy International at §§120-121, 131. So, as Lord Carnwath explained in Privacy International at §130:

the courts have … felt free to adapt or limit the scope and form or judicial review, so as to ensure respect on the one hand for the particular statutory context and the inferred intention of the legislature, and on the other for the fundamental principles of the rule of law, and to find an appropriate balance between the two.

Section 28

12. The key provision in the statutory overlay is s.29(3) of the Senior Courts Act 1981. But it is right to start one step back, with s.28(1) and (2)(a) of that Act. These deal with an appeal by case stated. They provide as follows:

(1) Subject to subsection (2), any order, judgment or other decision of the Crown Court may be questioned by any party to the proceedings, on the ground that it is wrong in law or is in excess of jurisdiction, by applying to the Crown Court to have a case stated by that court for the opinion of the High Court. (2) Subsection (1) shall not apply to – (a) a judgment or other decision of the Crown Court relating to trial on indictment …

In the absence of s.28(2)(a), the crown court's exercise of the Regulation 3 function would be amenable to appeal to the High Court by case stated. That is the position in the parallel situation of a case stated appeal pursuant to section 111 of the Magistrates Courts Act 1980, where a magistrates' court exercises the Regulation 3(1)(a) function: see eg. Coll v DPP [2022] EWHC 2635 (Admin). On a case stated appeal the High Court avoids becoming embroiled in arguments about what Parliament meant by "wrong in law" and "in excess of jurisdiction". That is because, as has been seen (Taylor at §40), "wrong in law" embraces all material public law errors. But section 28(2)(a) means that case stated appeal is not available to challenge a crown court judgment or decision "relating to trial on indictment". That includes a Regulation 3 costs decision at the end of a crown court trial: see eg. Hunter v Crown Court at Newcastle upon Tyne [2013] EWHC 191 (Admin) [2014] QB 94.

Section 29(3)

13. Section 29(3) of the 1981 Act provides as follows:

In relation to the jurisdiction of the Crown Court, other than its jurisdiction in matters relating to trial on indictment, the High Court shall have all such jurisdiction to make mandatory, prohibiting or quashing orders as the High Court possesses in relation to the jurisdiction of an inferior court.

Matters Relating to Trial on Indictment

14. The meaning of "matters relating to trial on indictment" in these provisions has been addressed in a series of cases over the years, including at the highest judicial altitude of the House of Lords in the 1980s and 1990s: In Re Smalley [1985] 1 AC 622; In Re Sampson [1987] 1 WLR 194; R v Manchester Crown Court, ex p Director of Public Prosecutions [1993] 1 WLR 1524; and In Re Ashton [1994] 1 AC 9. As I have already mentioned, in three of the High Court cases in the Harrow (Maidstone) line of authority, the impugned decision in the judicial review claim was a decision pursuant to Regulation 3 as to costs. Those cases are CCE (Leicester); DPP (Sheffield) and DPP (Aylesbury). A fourth case, TM Eye, was a case about a decision refusing costs from central funds. All of these judicial review cases treat such decisions as falling within the s.29(3) phrase "the Crown Court … jurisdiction in matters relating to trial on indictment". That fits with the approach to "a judgment or other decision of the Crown Court relating to trial on indictment" in s.28(2)(a), as seen in the case stated appeal case of Hunter.

The Harrow (Maidstone) Cases

15. Beginning with R v Maidstone Crown Court, ex p Harrow London Borough Council [2000] QB 719, there has been a developing line of High Court cases which identifies a judicial review jurisdiction of the Administrative Court, applicable notwithstanding that the impugned order, judgment or decision of the Crown Court is assessed as falling within "matters relating to trial on indictment" in s.29(3). Reference is made to this line of cases in R (Belhaj) v DPP [2018] UKSC 33 [2019] AC 593 at §16. The key cases in this line of authority which were provided to me, and addressed by Counsel, were these: Harrow (Maidstone) (30.4.99); R (Commissioners of Customs and Excise) v Crown Court at Leicester [2001] EWHC Admin 33 (24.1.01); R (Kenneally) v Snaresbrook Crown Court [2001] EWHC Admin 968 [2002] QB 1169 (27.11.01); R (B) v Crown Court at Stafford [2006] EWHC 1645 (Admin) [2007] 1 WLR 1524 (4.7.06); R (Crown Prosecution Service) v Crown Court at Guildford [2007] EWHC 1798 (Admin) [2007] 1 WLR 2886 (4.7.07) R (Faithfull) v Crown Court at Ipswich [2007] EWHC 2763 (Admin) [2008] 1 WLR 1636 (26.10.07); R (Director of Public Prosecutions) v Crown Court at Sheffield [2014] EWHC 2014 (Admin) [2014] 1 WLR 4639 (20.6.14); R (M) v Crown Court at Kingston Upon Thames [2014] EWHC 2702 (Admin) [2016] 1 WLR 1685; R (Director of Public Prosecutions) v Aylesbury Crown Court [2017] EWHC 2987 (Admin) [2018] 4 WLR 3 (24.11.17); and R (TM Eye Ltd) v Crown Court at Southampton [2021] EWHC 2624 [2022] 1 WLR 1114 (30.9.21).

The Two Core Purposes of Section 29(3)

16. Two things have consistently been identified as being Core Purposes of the provision made by Parliament in s.29(3). (1) The first Core Purpose is protecting the crown court trial process from interference by interlocutory applications for judicial review while such a trial is ongoing. (2) The second Core Purpose is that the verdict or sentence as the outcome of a crown court trial are exclusively governed by deliberately-designed mechanisms for appeal (and Attorney General's reference).
17. These two Core Purposes of s.29(3) are linked by a unifying idea: that material flaws within the course or outcome of a crown court trial fall to be addressed by the appeal (and reference) mechanisms which have deliberately been conferred on defence and prosecution in relation to conviction and sentence. That unifying idea was described by Mitchell J in Harrow (Maidstone) at 731B:

The rationale of the exclusion (whatever its true scope) is that 'matters' relating to trial on indictment are ordinarily amenable to appeal following conviction and to permit interlocutory-type appeals might seriously delay the trial.

18. The first Core Purpose (protecting crown court trial process from interlocutory interference) is reflected in Mitchell J's phrase: "to permit interlocutory-type appeals might seriously delay the trial". This Core Purpose was described as follows by Lord Woolf CJ in CCE (Leicester) at §29:

The primary purpose of section 29(3) is to avoid the criminal trial process being interfered with by applications for judicial review.

It was also described by May LJ in B (Stafford), in explaining why judicial review in that case did not impinge on the Core Purposes of s.29(3) (at §14):

Speaking generally, this limitation [ie. s.29(3)] is designed to prevent trials on indictment being delayed by challenges in the nature of interlocutory appeals. If the Crown Court makes an error and the defendant is convicted, he can appeal after conviction to the Court of Appeal (Criminal Division). The present claim will have no such effect.

19. The second key purpose (verdict and sentence exclusively governed by appeal and referral rights) is reflected in Mitchell J's phrase "ordinarily amenable to appeal following conviction". This Core Purpose was identified by Lord Phillips in CPS (Guildford) at §§13-14 and §16; and by Gibbs J in Faithfull at §§34 and 36. The underpinning of those passages is Lord Bridge's description in Smalley at 642E of s.29(3) as serving to:

exclude … judicial review in relation to the verdict given or sentence passed at the conclusion of a trial on indictment, both of which are subject to appeal …

Then, this description, from Lord Bridge again in Sampson at 196F:

certain orders made at the conclusion of a trial on indictment are excluded from judicial review as "relating to trial on indictment" not because they affect the conduct of the trial, but rather because they are themselves an integral part of the trial process. This is obviously true of the verdict and sentence…

The Harrow (Maidstone) Cases Analysed

20. In Harrow (Maidstone) in April 1999 the judicial review court granted the claim and set aside – evidently by making a quashing order – a supervision order made by the Maidstone crown court pursuant to s.5 of the Criminal Procedure (Insanity) Act 1964. This was a matter relating to trial on indictment (see 742A). The making of an appropriate order was a statutory duty pursuant to s.5, but Parliament had given the crown court alternative options from which the appropriate order was to be chosen. There was a statutory precondition to the making of a supervision order. No such order could be made unless (s.5(1)(a)), following a trial of the issue, the jury had returned a special verdict of not guilty by reason of insanity (725H). It was plain that there had been no trial of an issue and no special verdict. On any view, the statutory precondition to the exercise of the power was absent. The judicial review court had a residual jurisdiction where the crown court acted without "jurisdiction" (742B-C, 743C). I observe that this was a case about a statutory precondition which could not lawfully have been found to have been fulfilled. It was a clear-cut case. The crown court had failed to ask the legally required question: was there a special verdict? Any positive answer to that question would have been vitiated by material public law error. It did not matter whether the public law error was characterised as an error as to precedent fact, an unreasonable decision or a conclusion unsupported by any evidence.
21. I think what Harrow (Maidstone) decided was this. A crown court decision, albeit in a matter relating to trial on indictment, is susceptible to quashing on judicial review where the crown court's decision that a statutory precondition was satisfied is flawed by public law error.
22. In CCE (Leicester) in January 2001 the judicial review court refused a challenge to a costs order made by the Leicester crown court pursuant to Regulation 3 (the same power as in the present case). This was a matter relating to trial on indictment. Several of the Components of Regulation 3 (§4 above) were in play. There was an identifiable failure by the crown court judge to ask the legally required questions (see §§16 and 38). However, the Causation Precondition and Specification Duty were matters which were going to be addressed subsequently in the crown court (see §§25 and 31-32). That left the Default Precondition, an evaluative question. As to that, the crown court's decision – whether "right" or "wrong" – was one made within "jurisdiction" (§§26-27) and with the necessary "material" (§27). In my judgment, this case was ultimately about a merits evaluation as to whether a statutory precondition was met, and the absence of a substitutionary review. The reference to "right" or "wrong" meant on the merits. This was not an unreasonable decision or a conclusion without evidence; hence the reference to the necessary "material".
23. I think what CCE (Leicester) decided was this. A crown court decision, albeit in a matter relating to trial on indictment, is not susceptible to quashing on judicial review where the crown court's decision that a statutory precondition was satisfied is not flawed by public law error.
24. In Kenneally in November 2001 the judicial review court quashed a hospital order made by the Snaresbrook crown court pursuant to s.51(5) of the Mental Health Act 1983. This was (by majority 2-1) a matter relating to trial on indictment. So far as relevant, a statutory precondition to the exercise of the power was whether "it appears to the court… that it is… inappropriate to bring the detainee before the court" (see §3). The crown court judge had misinterpreted "inappropriate" as including 'inappropriateness to stand trial'. That error attracted the judicial review jurisdiction as a "misdirection" of sufficient "gravity" and sufficiently "fundamental" to amount to a "jurisdictional" error (§§38, 40, 43, 50). Pill LJ said that he expected the scope of the Harrow (Maidstone) jurisdiction to be the subject of further consideration in future cases (§40). The Court did not rule on Treasury Counsel's submissions that an unreasonable decision – whether relating to a statutory precondition or otherwise – would not attract judicial review (1173F).
25. I think what Kenneally decided was this. A crown court decision, albeit in a matter relating to trial on indictment, is susceptible to quashing on judicial review where the crown court's decision that a statutory precondition was satisfied is flawed by misdirection in law.
26. In B (Stafford) in July 2006 the judicial review court made declarations that, at a trial on indictment in the Stafford crown court: (a) a prosecution witness had been entitled to make informed representations on a defence application for disclosure of her psychiatric records; and (b) the trial judge had acted unlawfully in granting the defence application without securing that course. The Court held (§14) that there was jurisdiction, to grant declarations as to the rights of a non-party to a trial on indictment. The court emphasised that the Core Purposes of s.29(3) were intact, because the trial on indictment was not being delayed by an interlocutory challenge. The Court left open (§15): whether the case would have fallen within Harrow (Maidstone) to quash a crown court order made without jurisdiction; and whether a judicial review jurisdiction arose by interpreting ss.29(3) and s.9(2) of the Human Rights Act 1998 to avoid "an unsatisfactory human rights lacuna" (§15).
27. I think what B (Stafford) decided was this. A crown court decision, albeit in a matter relating to trial on indictment, is susceptible to judicial review where the claim, and the remedies, do not offend the Core Purposes of s.29(3).
28. In CPS (Guildford) in July 2007 the judicial review court declined the prosecution's claim for a quashing order in respect of an extended sentence imposed after crown court conviction. That was even though the sentence was not a lawfully available sentencing option and the crown court should instead have imposed an indeterminate sentence of imprisonment for public protection. The Court identified a Core Purpose of s.29(3): to prevent judicial review to challenge a "verdict or sentence" at the conclusion of a trial on indictment (§§13-16), including a sentence which exceeds the "jurisdiction" of the sentencing court (§16). Harrow (Maidstone) and Kenneally did not govern that situation (§§14-16). It had been addressed in the House of Lords cases (§§13-14) identifying "verdict and sentence" at the end of a trial on indictment as not being susceptible to judicial review (§19 above). The sole route to challenge a sentence was appeal (or Attorney General's reference) to the Court of Appeal Criminal Division (§17).
29. I think what CPS (Guildford) decided was this. A crown court conviction or sentence after trial, as a matter relating to trial on indictment, is never susceptible to judicial review, because it would offend the Core Purposes of s.29(3).
30. In Faithfull in October 2007 the judicial review court declined to make a quashing order or a declaration (§§11, 21), claimed by a non-party to the crown court trial, in respect of the Ipswich crown court judge's legally erroneous failure (§23) to add a 'back-to-back' compensation order to a confiscation order. Judicial review did not lie under the Harrow (Maidstone) line of authority on jurisdictional error (§20); nor via the B (Stafford) route of granting a declaration at the suit of a non-party (§§21, 36). The challenge was squarely to part of "the sentence of the court" (§§27, 39), exclusively governed by rights of appeal deliberately not afforded to non-parties (§§29-30, 40), and a declaration would have a direct effect on sentence (§§36, 41). A Core Purpose would have been offended. CPS (Guildford) was applied. The Court did not think (obiter, as I read it) that the "error" was one as to "jurisdiction" (§35). As for a 'human rights lacuna' argument (§22), the availability of a compensation claim in the county court meant there was no lacuna (§§37, 42).
31. I think what Faithfull decided is the same was what CPS (Guildford) had decided. A crown court conviction or sentence after trial, as a matter relating to trial on indictment, is never susceptible to judicial review, because it would offend the Core Purposes of s.29(3).
32. In DPP (Sheffield) in June 2014 the judicial review court quashed a costs order made by the Sheffield crown court pursuant to Regulation 3 (the same power as in the present case). This was a matter relating to trial on indictment. The crown court judge's conclusion that the Default Precondition was satisfied involved a legally erroneous and legally impermissible approach. The judge had identified as an "improper act or omission" a prosecutorial decision not to prosecute others. The statutory power could not be exercised for that purpose ("as a means of impugning the prosecutorial discretion"), which meant the decision was "outwith the statutory power" (§16) and made without "jurisdiction" in the Harrow (Maidstone) sense (§23). This was not a case engaging the Core Purposes. It was not about a crown court conviction or sentence after trial; nor a claim which would involve a trial on indictment being delayed by an interlocutory challenge. I observe that this characterisation of statutory power being exercised other than for the purpose for which it had been conferred could equally have been characterised as another species of public law error: misdirection in law as to the meaning of "improper act or omission" in the statutory precondition (the Default Precondition). That would fit with what Kenneally decided. The Court left open whether any failure to apply "the right test", or a failure in not "hearing argument" (the Procedural Fairness Guarantee: §4 above), would also be public law errors attracting the judicial review jurisdiction (see §27).
33. I think what DPP (Sheffield) decided was this. A crown court decision, albeit in a matter relating to trial on indictment, but not offending the Core Purposes of s.29(3), is susceptible to judicial review for exercising the power other than for the purpose for which it was conferred.
34. In M (Kingston) in July 2014 the judicial review court quashed an order made by the Kingston crown court pursuant to s.35 of the Mental Health Act 1983, remanding an accused person to a specified hospital for a mental health report (§§21-22). This was a matter relating to trial on indictment (§31). It was not a case engaging the Core Purposes. It was not about a crown court conviction or sentence after trial; nor a claim which would involve a trial on indictment being delayed by an interlocutory challenge. The crown court judge had interpreted s.35 as permitting an order to be made for the purpose of obtaining evidence relevant to an issue at trial but, on its legally correct interpretation, the power could not lawfully be used for that purpose (§23). This was not a proper purpose (§29). There was a defect as to "jurisdiction" which was sufficiently "severe" and of sufficient "gravity" to attract the judicial review jurisdiction (§32). I observe that the public law error could equally have been characterised as a misdirection in law; but that, however the public law error was characterised, it did not relate to a statutory precondition.
35. I think what M (Kingston) decided was the same as DPP (Sheffield) decided. A crown court decision, albeit in a matter relating to trial on indictment, but not offending the Core Purposes of s.29(3), is susceptible to judicial review for exercising the power other than for the purpose for which it was conferred.
36. In DPP (Aylesbury) in November 2017 the judicial review court quashed an order made by the Aylesbury crown court pursuant to Regulation 3 (§5) (the the same power as in the present case). This was a matter relating to trial on indictment. It was not a case engaging the Core Purposes. It was not about a crown court conviction or sentence after trial; nor a claim which would involve a trial on indictment being delayed by an interlocutory challenge. The crown court judge had made public law errors in relation to three of the Components (§4 above) of Regulation 3. First, as to the interpretation of the phrase "on behalf of" in the Default Precondition. As to that, the judge had erroneously interpreted "on behalf of" as apply to the acts of an expert called by the prosecution. That was an error as a matter of legal analysis (§20). In other words, a misdirection in law. Secondly, as to the approach to "improper", also in the Default Precondition. As to that, the judge had made multiple public law errors: the failure to make a required finding of impropriety by the CPS (§§24, 30); the absence of any "proper basis" or any evidence for such a conclusion (§§25, 29); and insufficiency of enquiry (§30). Thirdly as to the Specification Duty (Regulation 3(3)). As to that, the judge had failed to discharge this duty, which meant that the order was ultra vires (§32). The Court concluded that it had jurisdiction to deal with these "various flaws in the judge's approach", which were flaws of sufficient "gravity" to amount to "jurisdictional errors" in the Harrow (Maidstone) sense. I observe that misdirection in law, failure to make a required finding, a conclusion unsupported by any evidence, insufficiency of enquiry and failure to discharge a duty are all classic species of public law error. I also observe that whereas most of the public law errors in DPP Aylesbury were in the approach to a statutory precondition (the Default Precondition), the breach of the Specification Duty was not.
37. I think what DPP (Aylesbury) decided was this. A crown court decision, albeit in a matter relating to trial on indictment, but not offending the Core Purposes of s.29(3), is susceptible to judicial review where the crown court's decision is flawed by public law error.
38. In TM Eye in September 2021 the judicial review court refused to quash (§2) the Southampton crown court judge's refusal to make an order for costs out of central funds in favour of a private prosecutor who had secured a conviction. This was a matter relating to trial on indictment. It was not a case engaging the Core Purposes. It was not about a crown court conviction or sentence after trial; nor a claim which would involve a trial on indictment being delayed by an interlocutory challenge. In relying on the wealth of the private prosecutor, the judge had based his decision on an "irrelevant consideration". That was an error of sufficient "gravity", to take the decision outside the crown court's "jurisdiction", to constitute a basis for intervention within the jurisdiction of judicial review. The judicial review court would have intervened, except for factors which justified the refusal of a remedy. These were the following: that the private prosecutor had itself contributed to the judge's error by not identifying and addressing the correct legal principles; the judge had declined for legitimate reasons to reconsider the refusal; and the outcome would not have been substantially different had the judge's error not been made. I observe that a decision made in reliance on a legally irrelevant consideration is a classic public law error.
39. I think what TM Eye decided was this. A crown court decision, albeit in a matter relating to trial on indictment, but not offending the Core Purposes of s.29(3), is susceptible to judicial review where the crown court's decision is flawed by public law error.

"Gravity", "severity" and "jurisdiction"

40. Throughout the Harrow (Maidstone) line of cases there are references to "gravity", "severity" and "jurisdiction". It is necessary to grapple with these. If we start with "gravity" and "severity", the question is whether these are intended to be an added feature, beyond the vitiating error being an error going to "jurisdiction". Logically, there are two possibilities. The first is that there is no added feature. The Courts are speaking of errors and flaws which are "grave" and "severe" in the sense that they go to "jurisdiction". In other words, the crown court has not simply made a mistake but has made a "grave" or "severe" mistake because it is a mistake going to "jurisdiction". The second possibility is that there is an added feature. The Courts are starting with errors and flaws which go to "jurisdiction", and then speaking of a narrowing subspecies of those: errors and flaws which go to "jurisdiction", and which are also "grave" or "severe".
41. In my judgment, there is no added feature. There is no narrowing sub-species. It is enough that the error is serious (or "grave" or "severe") in the sense that it goes beyond being "right" or "wrong" (see CCE (Leicester)). It is beyond a substitutionary merits disagreement on an evaluative judgment. It is enough that it is an error going to "jurisdiction". I am fortified in that conclusion by two things. The first is that the Harrow (Maidstone) cases nowhere identify any further standard of gravity or severity. There is no Cart-style formulation. If such a standard existed, it would be extremely odd if none of the cases attempted to articulate what it was. The quality of law in this area would be materially undermined if there were some elusive superadded standard, being required by the courts, but never being explained by the courts. The second thing is that this idea – seriousness and severity – has clearly been treated as synonymous with error as to "jurisdiction" not additional to it. Thus there is one concept; there are not two. The search is for an error as to "jurisdiction". In DPP (Aylesbury), Sharp LJ (as she then was) drew specific attention (at §8) to a passage from M (Kingston) at §32, describing Kenneally as a binding decision to the effect that, where an order is made relating to a trial on indictment, it may nevertheless be quashed in circumstances where "the defect is so severe that it deprived the court below of jurisdiction to make it"; and so the question is whether there is a jurisdictional error "of such gravity as to take the case out of the jurisdiction of the crown court". That is why DPP (Sheffield) spoke (at §23) of Harrow (Maidstone) as recognising jurisdiction if a decision was "made without jurisdiction" (see DPP (Aylesbury) §9). This is why judicial review was granted in DPP (Aylesbury) because the "various flaws" were "of sufficient gravity to amount to jurisdictional errors" (see §35). The search is for a "jurisdictional error": error as to "jurisdiction". Seriousness and severity are part of that description; not some elusive superadded component.
42. The focus is therefore on what "jurisdictional error" actually means, in the context of the Harrow (Maidstone) cases. One clear possibility would be that the courts have been deploying the familiar concept, known to the law, of "jurisdictional error in the narrow, pre-Anisminic sense" (Cart (SC) at §34), or "pre-Anisminic excess of jurisdiction" (at §38). The history and the nature of that concept is discussed in the judgments in Cart (DC) and Cart (SC).
43. In my judgment, the Harrow (Maidstone) cases are not a pocket of the current law which have returned to this "pre-Anisminic" error going to "jurisdiction". There are three main reasons for that conclusion. The first is that I have been unable to find any indication in any of the Harrow (Maidstone) cases that this is what the judges who decided those cases meant. They would surely have said so, had they been reverting to pre-Anisminic distinctions. The second reason is that such a return would be extremely surprising. It would have been a return to the "many technicalities" of the past (Cart (SC) §40), involving artificiality and technicality (Cart (SC) §111). It would adopt an approach identified as objectionable in principle, lacking in justification and not promoting the rule of law (Cart (SC) §110; Privacy International §84). I cannot accept that the Courts in the Harrow (Maidstone) line of cases intended to perpetuate this discredited artificial distinction. The third reason is the proof of the pudding. The sorts of errors of approach seen in the cases as errors going to "jurisdiction" are conventional public law errors, which is why I listed them (§8 above).

Three Possible Bases of Jurisdiction

44. As I see it, there are three distinct analytical bases on which the judicial review court could conclude that there is jurisdiction in the present case. Other formulations are available, but in the present case I see these as the three candidates that really matter. First, there is what I will call "the Expansive Basis". This – as I see it – would involve the following proposition:

The Expansive Basis. A crown court decision concerning a matter relating to trial on indictment is susceptible to judicial review for public law error, even if it would offend the Core Purposes of s.29(3) (by delaying a trial by interlocutory challenge or impugning a verdict or sentence after trial).

45. Secondly, there is what I will call "the Narrow Basis". This – as I see it – would involve the following proposition:

The Narrow Basis. A crown court decision concerning a matter relating to trial on indictment is susceptible to judicial review for public law error in the crown court's application of a statutory precondition to a power or duty, provided that it would not offend the Core Purposes of s.29(3) (by delaying a trial by interlocutory challenge or impugning a verdict or sentence after trial).

46. Thirdly, and occupying a space between these two, there is what I will call "the Intermediate Basis". This – as I see it – would involve the following proposition:

The Intermediate Basis. A crown court decision concerning a matter relating to trial on indictment is susceptible to judicial review for public law error, provided that it would not offend the Core Purposes of s.29(3) (by delaying a trial by interlocutory challenge or impugning a verdict or sentence after trial).

The Expansive Basis

47. Mr Molyneaux mounts an impressive and sustained argument in favour of the Expansive Basis (§44 above). Its essential elements, as I saw it, are as follows. (1) The previous line of House of Lords authorities, which would allow unchecked public law error in cases concerning a matter relating to trial on indictment through a perceived s.29(3) exclusion, cannot stand in light of the subsequent Supreme Court authorities of Cart and Privacy International. (2) Although designated as a superior court of record, the crown court remains a court of limited jurisdiction (see eg. R v Beck [2003] EWCA Crim 2198 at §27; R (Chaudhary) v Bristol Crown Court (No.2) [2014] EWHC 2014 (Admin) [2014] 1 WLR 4639 at §36), and not the 'alter ego' of the High Court (Privacy International §160). (3) Given that Parliament has conferred statutory powers and imposed statutory duties, there is a recognised constitutional principle that judicial review cannot be regarded as having been dispensed with (Cart (DC) §§36-40; Cart (SC) §§30, 37; Privacy International at §§116, 160). This is no more than the traditional approach to 'no certiorari' provisions (Cart (DC) §83; Privacy International §§72, 123). (4) There is a statutory construction which provides the principled answer. We need to stop focusing on "matters relating to trial on indictment". Section 29(3) refers to "the Crown Court" in "its jurisdiction" in such matters. The word "jurisdiction" means "jurisdiction (exercised lawfully)"; not "jurisdiction (even if exercised unlawfully)". This is directly parallel to the principled interpretative solution which the historic case of Anisminic itself adopted as regards the word "determination" (Privacy International §§2, 54). (5) This straightforward approach, and the principled interpretation of "jurisdiction", best explains how the Harrow (Maidstone) cases have come to recognise a judicial review jurisdiction notwithstanding the wording of s.29(3). It avoids technical and elusive distinctions, relying instead on (a) conventional public law error and (b) discretionary bars to judicial review.
48. I cannot accept the Expansive Basis. That is for these reasons. (1) It produces a scope for judicial review which gives no recognition to Parliament's Core Purposes in enacting s.29(3). It assumes an all-or-nothing role for judicial review, an assumption rejected in Cart itself. It would be flatly inconsistent with the application in the Harrow (Maidstone) authorities of the first Core Purpose (see eg. B (Stafford)) and the second Core Purpose (see CPS (Guildford) and Faithfull). It would mean concluding that CPS (Guildford) and Faithfull were wrongly decided. It would mean decades of cases about s.29(3) are wrong, including the original House of Lords cases about s.29(3). It would allow judicial review against conviction or sentence, where there is no right of appeal (or reference) and therefore no alternative remedy. It would allow a victim – as a person with a sufficient interest – to challenge a sentence by judicial review. It would allow a judicial review challenge to a conviction, or an acquittal. (3) The argument deletes "other than its jurisdiction in matters relating to trial on indictment" from s.29(3). It would mean every decision is susceptible to challenge in the High Court for public law error. It would mean this is a statutory scheme where there is a public law error jurisdiction (appeal by case stated where the decision is "wrong in law or is in excess of jurisdiction") against crown court decisions with an exception for matters relating to trial on indictment (s.28(1)(2)(a)), and then a public law error jurisdiction (judicial review) including everything within that exception (s.29(3)). (3) Such a course is not, in my judgment, open to a High Court judge nor – as I see it – a Divisional Court. But, given the potency of the points about the lack of recognition given to Parliament's Core Purposes in enacting s.29(3) and the all-or-nothing role for judicial review, I would not adopt this analysis even were it open to me. I can see that the Courts could say (a) the judicial review jurisdiction is indeed all-encompassing but (b) the Core Purposes are protected by discretionary bars such as prematurity and alternative remedies. The judicial review court could address questions about protecting crown court trials from satellite interlocutory challenge, just as it does for magistrates' court proceedings: see eg. R (DPP) v Walsall Magistrates' Court [2019] EWHC 3317 (Admin) [2020] ACD 21 at §46). But once the position is reached that, for the crown court, the Core Purposes require strong protection, the position is in substance the Intermediate Basis.

The Narrow Basis

49. The Narrow Basis (§45 above) would be sufficient for the judicial review court to have jurisdiction in the present case. The various ways in which this claim is put involve public law error in the crown court's decision regarding a statutory precondition to a power: the Causation Precondition. The claim does not offend the Core Purposes of s.29(3): the claim would neither delay a trial by interlocutory challenge nor impugn a verdict or sentence after trial. Mr Molyneaux submits, and Mr Puzey does not contest, that there is jurisdiction, for a public law error (or "error of law") in the application of the Causation Precondition to the exercise of the Regulation 3 power.
50. I agree. I accept that the judicial review jurisdiction extends at least this far. My reasons are straightforward. (1) This analysis accommodates the Core Purposes of s.29(3). In that respect, it fits with B (Stafford), CPS (Guildford) and Faithfull. (2) This analysis fits with Harrow (Maidstone); Kenneally; DPP (Sheffield); and DPP (Aylesbury). These cases convincingly illustrate material public law errors in a decision as to a statutory precondition. (3) No case in the line of Harrow (Maidstone) cases conflicts with this analysis. In CCE (Leicester) there was no material public law error as to a statutory precondition. (4) Once it is recognised, in a case not engaging the Core Purposes, that a public law error in the application of a statutory precondition is susceptible to judicial review, I can see no justifiable basis for a 'pick and mix' approach to which public law errors (in the application of a statutory precondition) would qualify and which would not. None is given in the authorities.

The Intermediate Basis

51. My conclusion on the Narrow Basis is sufficient for me to be satisfied that there is jurisdiction in the present case. But I have analysed the Expansive Basis and I will say what I made of the Intermediate Basis (§46 above).
52. There are undoubtedly reasons why the judicial review court's jurisdiction could stop at the Narrow Basis. Among them are these. The Narrow Basis focuses clearly on the application of statutory preconditions. That is not a vague or elusive scope. It is clear and workable. It also has the advantage of explaining why there are the repeated references to "jurisdiction" in the Harrow (Maidstone) line of cases. It explains the ideas of "gravity" and "severity". A public law error in the application of a statutory precondition means the crown court never got to the heart of the power or duty, because it went off the rails in applying prior questions. So, the word "jurisdiction" refers to the crown court's powers or duties, once a lawful decision (free from public law error) has been made as to its statutory preconditions. This approach can claim virtues, in an exercise of statutory interpretation, in explaining the words used by Parliament. In s.29(3), the phrase "other than its jurisdiction in matters relating to trial on indictment" is – in effect – being interpreted to mean "other than its jurisdiction (with statutory preconditions lawfully applied) in matters relating to trial on indictment". That would also enable a distinct meaning to be given to the otherwise otiose phrase "in excess of jurisdiction" in s.28(2)(a), namely: "with a statutory precondition unlawfully applied". Finally, the Narrow Basis has the virtue of being firmly linked to the source authorities in the Harrow (Maidstone) line, especially Harrow (Maidstone) itself and Kenneally.
53. I do not think the judicial review court's jurisdiction stops at the Narrow Basis. In my judgment, the Intermediate Basis is correct in law. My reasons are as follows. (1) The Intermediate Basis recognises the Core Purposes and draws the line, clearly, at preventing them from being undermined. (2) This analysis achieves Lord Carnwath's imperatives (§11 above) of ensuring respect on the one hand for the particular statutory context and the inferred intention of the legislature, and on the other for the fundamental principles of the rule of law, finding the appropriate balance between the two. It reflects the constitutional touchstone (§11 above): the need to secure the level of scrutiny required by the rule of law. (3) This analysis alone explains the authorities in the Harrow (Maidstone) cases. It explains B (Stafford), a case which did not concern a statutory precondition to a power or duty, and where susceptibility to judicial review was explained by the fact that the Core Purposes were protected. It explains why DPP (Sheffield) – a case about a statutory precondition – was articulated more broadly: that the statutory power being exercised other than for the purpose for which it had been conferred. It alone explains M (Kingston) – a case not about a statutory precondition – in which the public law error was that the statutory power was being exercised other than for the purpose for which it was conferred. It alone explains why in DPP (Aylesbury) breach of the Specification Duty – not a statutory precondition to the power – was one of the vitiating features for the purposes of judicial review. It also fits with TM Eye, which was a straightforward public law error. (4) Although the Narrow Basis is clear and workable, it is not just or satisfactory. Assuming that the Core Purposes are protected, why should a material public law error vitiate the crown court decision when it relates to a statutory precondition, but not when it relates to the exercise of power or discharge of duty? Why should procedural unfairness vitiate a decision only when it can be said to be a public law error relating to a statutory precondition? This can be illustrated by reference to Regulation 3. The Procedural Fairness Guarantee ("after hearing the parties": §4 above) can be characterised as a statutory precondition. But common law procedural fairness would apply in the discharge of the statutory function. Why should unreasonableness, or reliance on a legal irrelevancy, or the giving of legally inadequate reasons, or a legal insufficiency of enquiry, vitiate a decision only when they relate to a statutory precondition? Down the path of the Narrow Basis would inevitably lie two things. First, the shoehorning of public law errors into 'preconditions', whether statutory preconditions or public law preconditions. Secondly, the relaxation of the idea of 'preconditions', statutory or otherwise. Which brings us full circle, because these were the very tensions which made the pre-Anisminic law of "jurisdictional" error so unsatisfactory. Meanwhile, the price paid is uncertainty and unpredictability.
54. There are other points too. First, that the Narrow Basis does not in fact provide a neat statutory interpretation of "other than its jurisdiction in matters relating to trial on indictment". It does not mean "other than its jurisdiction (with statutory preconditions lawfully applied) in matters relating to trial on indictment". This formulation does not account for protection of the Core Purposes. It would mean an interlocutory judicial review could interfere with the trial process. It would mean a verdict or sentence could be challenged. So, there is no neat interpretative solution. Secondly, that there is little utility searching for the correct interpretation of "in excess of jurisdiction" in s.28(2)(a). On any view, that phrase it otiose, because it is swallowed up by its twin: "wrong in law". Thirdly, ever since Anisminic and the collapse in public law of the distinction between "jurisdictional error" and "public law error", they have been different ways of saying the same thing (Privacy International §§43, 153). Each involves "gravity" and "severity". We can read the cases – as they have developed – as meaning there was ultimately no confining magic in the use of the word "jurisdiction". As Professor Hungerford-Welch pointed out in his Criminal Law Review commentary at [2018] Crim LR 333, the Court's approach in DPP (Aylesbury) is "entirely consistent" with the Anisminic case, where public law error (such as "misconstruction of the relevant law") makes the decision a "nullity" (as to which, see Privacy International §107). Fourthly, the Intermediate Basis removes the need for arguments based on: (i) declaratory relief as the remedy sought, or (ii) a special basis for inclusion of HRA-based grounds. Such arguments had featured in B (Stafford) and in Faithfull. They featured, as fallback arguments, in the present case.

Part 3: Application

What Happened at the Crown Court

55. I have satisfied myself on the legally prior issue of jurisdiction. There is no question of undermining the Core Purposes of s.29(3). If it matters, the claimed public law error relates to a statutory precondition: the Causation Precondition. I turn to whether there was a public law error. The context and circumstances were as follows. In May 2022 Exolum was being prosecuted by the HSE in the Great Grimsby Crown Court. The charges were failing to discharge the duty owed to employees, and the duty owed to non-employees, in contravention of ss. 2 and 3 of the Health and Safety at Work 1974. The events in question related to March 2018 and arose in the context of a fuel pipeline. The trial on indictment was before HHJ Fanning ("the Judge") and a jury. It had a 10 day trial slot beginning on Monday 16 May 2022 (Day 1) and ending on Friday 27 May 2022 (Day 10). Exolum had earlier made an application to stay the indictment on the ground of abuse of process. This had been heard on 14 and 15 February 2022 by the Judge and dismissed in a ruling delivered by him on 15 February 2022. For the purposes of the issues in the present case, it is necessary but also sufficient to identify the following key points within the trial:
56. First, there was the position on Day 1. On that day, housekeeping matters were dealt with (starting at 12:22), before a jury was empanelled and sworn in. The Judge's opening remarks to the jury followed and the case adjourned at 15:26 for the HSE's opening speech to take place at the beginning of Day 2 (Tuesday 17 May 2022). Before the jury was empanelled, the Judge asked the advocates about timetable. Counsel for Exolum (Mr Cooper KC) referred to his previous description of 10 days as "ample" and said he anticipated that the jury would be sent out "probably" on Day 8 (Wednesday 25 May 2022).
57. Secondly, there was the position on Day 4 (Thursday 19 May 2022). On that day, in a break in the prosecution evidence there was a discussion (at 11:20) about the progress of the case. Mr Cooper KC told the Judge that having "taken stock on timings", it was expected to "get through the Crown's case today". There was to be a half-time application to dismiss the case with "the argument tomorrow morning" followed by a ruling which would follow, depending on how long the Judge needed ("it's a matter for Your Honour as to how long you need, if you wanted to reflect on it"). If the case was proceeding, and if mid/late afternoon had been reached, then the plan was to start the defence evidence on Monday (Day 6). The Judge asked, on that basis, "how are we doing in your opinions in terms of progress?" Mr Cooper KC said that, "trying to be as realistic as possible" and "building in possible court issues", he anticipated a discussion about route to verdict on Wednesday "probably late morning, early afternoon" with speeches "either late Wednesday or first thing Thursday". It was "the most likely" that the jury would be out on Friday morning (Day 10), but Thursday afternoon (Day 9) was "a possibility". This depended on the Judge: "It depends how long your summing up is". The Judge responded that he would alert the listing officer "that there may be an issue" about "going beyond next week", but there was no need to raise this with the jury because that would be to worry them "unnecessarily".
58. Thirdly, there was the position on Friday 20 May 2022 (Day 5). On that day (at 11:05) an issue was raised by the defence relating to the Official Secrets Act 1989. I will call this the "OSA Issue". In the light of that point, significant time was lost in the trial. There were hotly disputed questions about where responsibility lay for the late raising of the OSA Issue.
59. Fourthly, there was the position on Monday 23 May 2022 (Day 6). On that day the Judge heard submissions and gave a ruling (13:24) on whether the case could proceed, deciding that it could.
60. Fifthly, there was the position on Tuesday 24 May 2022 (Day 7). Mr Cooper KC told the Judge (11:44) that it was not possible to give the Judge assurance as to when the defence evidence might finish: "based on all the[] imponderables" arising out of the OSA Issue, this was not possible. The Judge made enquiries and (at 13:11) told the parties that "my estimate is there's six to eight days of this trial left, probably nearer the eight, if we have jury time to consider, which, of course, we do". The Judge explained that five jurors could not sit beyond Day 10. In those circumstances, the trial was abandoned. At 13:20 the Judge discharged the jury. At 14:33 there was a discussion about a new date. Mr Cooper KC said to the Judge: "I think it is wise to set aside three weeks, is it not, given what has happened?" The Judge agreed. The retrial was listed for 3 weeks.
61. Sixthly, there was the position on Wednesday 25 May 2022 (Day 8). On that day, the Judge heard submissions (12:28) from the advocates for the prosecution (Mr Puzey) and defence on the prosecution's application for a costs order pursuant to Regulation 3. In an ex tempore ruling given by the Judge at the end of the oral submissions, the application was granted. Exolum was ordered to pay HSE's costs in the sum of £69,650.18. That is the decision which is challenged in these judicial review proceedings. The central issues argued between the advocates overlapped. One key question was whether the failure by Exolum to raise the OSA Issue well ahead of trial was a failure which satisfied the Default Precondition. Another key question was whether it had been Exolum's failure, or rather the prosecution evidence, which had led to the abandonment of the trial. Mr Cooper KC maintained that the trial had been abandoned, not because of any default by Exolum, but because of evidence given by a prosecution witness in cross-examination. The Judge rejected that contention.

The Judge's Ruling

62. The Judge's ex tempore ruling granting HSE's Regulation 3 costs application occupies 7½ pages of the transcript. The passages which bear most directly on the Causation Precondition are as follows. The numbering is mine, for ease of cross-referencing later. I make clear that what follows here is an extract of selected passages, starting 3 pages into the 7½ page ruling:

[1] … But really, on the issues that were laid out on day one of this trial, the question really is: could the jury have heard the case, had the time to consider the evidence without an undue pressure of time to deliver verdicts within the ten days that this trial was calculated to last? And it seems to me on day one, that clearly was the position. Absolutely ten days was adequate, otherwise I would have been told to the contrary and I'd have canvassed enough jurors on a panel to sit into a third week.

[2] As it is, I've discharged the jury. Why is that? It's because only seven of them could sit beyond this Friday, and that's even having given consideration to my ranging of the remaining days of the trial over the two subsequent weeks. The reasons for the other five members of the jury that I've discharged being unable to sit on beyond Friday were, on my assessment of them, perfectly valid reasons and, in fact, insurmountable reasons. So we are, as a result of which I've said I've discharged the jury.

[3] Why then was it canvassed at all that we would need to sit beyond the ten days originally planned? Well, Mr Cooper QC contends that that is as a result of evidence given by Mr Pitman to him in the course of cross-examination. And his submission essentially is there's been such a change in the prosecution's position that, on taking further instructions, Mr Land – who is, in effect, the embodiment of the corporate defendant, so to speak – was concerned he would have to reveal information that he considers would put him in breach of the contract between his company and the MOD, and him specifically in breach of obligations imposed upon him, he said, by the Official Secrets Act 1989, and so concerned was he that he wanted an opportunity to obtain independent legal advice…

…

[4] … [W]e are where we are. It was impossible to continue the trial within the original window; and impossible, even though I tried, to expand that window – and the reason I can't expand it, as I've already said, is because of jury difficulties. So then the issue is: why are we here? It's simple, in my view: it is down to the abject failure of the defendant to give any consideration to any issues relating to its contract with the MOD, to give any consideration to the effect of the Official Secrets Act 1989 on Mr Land and upon others. And it cannot have been other than obvious to this company, and Mr Land, and its group legal director, that issues relating to the contract and the operation of the pipeline are at the heart of this trial.

[5] So why did the issue only arise almost at the conclusion of the prosecution evidence and midway through the evidence of Mr Pitman? Well, I'm told that it's a fundamental change on the part of the way the prosecution put its case that fundamentally altered the position of the defence. I do not accept that…

[6] So the days spent on this trial to date have been a complete and utter waste of time, and Mr Puzey, the prosecutor, applies for the costs incurred by the prosecution over the course of the last eight days be met by the defendant. He's set out all of this in accordance with the requirements of Criminal Procedure Rule 45.8. The relevant discretionary power he asks me to exercise, and which I've got to identify, is that in s.19 of the Prosecution of Offences Act 1985. The relevant test is set out in Archbold, it's in Blackstone's, it's in Archbold Magistrates' Court. In the main, the focus in those cases is on the position of an application against a prosecutor. This is a rather more novel situation: an application against a defendant in the circumstances that I have outlined already. But the test is the same, it's paraphrased in the Criminal Practice Direction at paragraphs 4.11 and 4.12. Firstly, has there been an unnecessary or improper act or omission? Secondly, as a result of such an unnecessary or improper act or omission, have costs been incurred by another party? And if the answer to both of those questions is "yes", then should the court exercise what is a discretion and order the party responsible to meet the whole, or any part, of the relevant costs and, if so, what is the specific sum involved?

[7] … [H]ad the defendant had regard – and the defendant, not just his legal team, must have regard to its obligations under the Criminal Procedure Rules, as set out in Mr Puzey's application, then this defendant would not be in the position that it is in today – that is facing this application in a stalled trial, and nor would the rest of us. So has there been an unnecessary or improper act or omission? Yes, I've identified it. As a result, have costs been incurred by another party? Yes, they have, and they've been set out and agreed; indeed, there's no disagreement as to quantum. Should I exercise my discretion? Well, this is not about punishing the defendant; it's about compensating the prosecutor for a loss that it could not have contemplated, and for which it was not responsible. And in answer to the simple question that I pose, "Should the HSE bear the costs of the defendant's failure in this case?" the answer is simply that it should not. And therefore, it follows that I do make the order that the defendant do bear the wasted costs as agreed between the parties, the quantum as agreed between the parties, and I think the figure is £69,650.18…

The Agreed Issue

63. The agreed issue which arises is this: Was the Judge's conclusion that the HSE incurred costs "as a result of" Exolum's failure to raise the OSA Issue before the trial vitiated by an error of law? This issue solely concerns the Judge's approach to the Causation Precondition. No challenge is maintained as to anything else, including the Judge's conclusions on the Default Precondition, the procedure (including the statutory requirement as to "hearing the parties"), or the Judge's exercise of discretionary power. The premise is that the Judge unassailably found that Exolum's failure to raise the OSA Issue before the trial was an "improper omission". The challenge relates to the finding that HSE's costs of the abandoned trial were incurred "as a result" of that default.
64. The parties identified a further substantive question, the answer to which – they agree – stands or falls with the agreed issue, and the answer to which is common ground. They are agreed that, if the Judge's conclusion that the HSE incurred costs "as a result of" Exolum's failure to raise the OSA Issue before the trial was vitiated by an error of law, then it would follow in law that the Judge's decision to impose costs breached Exolum's A1P1 (property) rights. The basis of this agreed position is that error of law would breach the A1P1 requirement of interference with property rights requiring to be as "provided for by law". Exolum cites Iatridis v Greece (2000) EHRR 97 at §58.

Exolum's Argument

65. Mr Molyneaux for Exolum (who did not appear below) submits that there was a material public law error in the Judge's conclusion that HSE incurred costs "as a result of" Exolum's failure to raise the OSA issue before the trial. I have already referred (§9 above) to the overlapping ways in which the claim is put. The essence of his argument – as I saw it – was as follows:
66. The Causation Precondition was a statutory condition and so a necessary step before the Judge could make any adverse decision in the exercise of the discretionary power to impose a regulation 3 costs order on Exolum. The Judge addressed the Default Condition and found it satisfied, identifying the relevant default. Having reached that conclusion, the Judge needed to address the Causation Precondition. He needed to be satisfied that costs incurred by HSE in respect of the proceedings had been incurred "as a result of" that improper omission. He needed to say why. His finding needed an evidenced basis and legally adequate reasons. It is true that the Causation Precondition posed a question for the Judge's evaluative judgment of facts. It was not a 'hard-edged' question of law or of precedent fact (or objective fact), whose incorrectness would be a vitiating error of law. But basic public law duties required the Judge to ask the right question, to conduct a legally sufficient enquiry, to reach an adverse conclusion only if supported by evidence, and only if reasonable, to adopt a legally correct approach, and to give legally adequate reasons. What was needed was a reasoned decision which "covered the correct ground and answered the right questions" (Anya v University of Oxford [2001] EWCA Civ 847 [2001] ICR 847 at §26), to "enable the parties … readily to analyse the reasoning that was essential to the judge's decision" (English v Emery Reimbold & Strick Ltd [2002] EWCA Civ 605 [2002] 1 WLR 2409 at §21). It was not enough that one part of the Ruling described the relevant legal test; the question is whether an objective reading of the Ruling as a whole shows that the Judge understood the test and applied it legally correctly (cf. R (AAA) v SSHD [2023] EWCA Civ 745 §§73, 129).
67. The Judge failed to discharge these basic public law duties, any one of which would be sufficient to constitute a vitiating flaw and an error of law. The Causation Precondition required the Judge to pose 'counterfactuals' and apply a 'balance of probabilities' test. Since the relevant default was the failure by Exolum to raise the OSA Issue before the trial, the counterfactual involved that step having been taken. The first 'counterfactual' question is whether, on the balance of probabilities, the trial slot would have been longer than the ten days. If not, the second 'counterfactual' question would arise. It is whether – on the balance of probabilities – the ten day trial slot would have been adequate. The Judge did not analyse either of these questions. No reasoning grappled with them. It is wrong to speculate about how they would have been answered. In fact, the evidence did not support an adverse conclusion. On the first counterfactual, it is speculative to say that more than a ten-day time estimate would have been chosen, still less on the balance of probabilities. As to the second counterfactual, a ten-day time slot could not have been adequate, still less on the balance of probabilities. That is because the ten-day slot was already inadequate. This is reflected in two aspects of the evidence. First, the Day 4 discussion and prognosis that the jury were most likely to be sent out on Day 10. Secondly, the Day 7 description of "six to eight days of this trial left". The trial was already likely to need to be abandoned in any event. That is because – as was known at the time of the Regulation 3 decision – the jury could not have sat beyond Day 10. Indeed, in those circumstances the only reasonable conclusion open to the Judge was to find in Exolum's favour in the application of the Causation Precondition.
68. The Judge's approach was clearly legally deficient. He focused at [2] and [4] on the position as at Day 7 when the trial could not be completed following the time lost by virtue of the OSA issue having been raised. He focused at [1] on the position as at Day 1, when the court and the parties had been satisfied that the case could be dealt with within the 2 week slot, and the jury was subsequently empanelled on that basis. Entirely missing from the judge's assessment was consideration of the position as at Day 4, prior to the OSA issue being raised. Also missing was recognition of what he had said on Day 7. The Judge failed to address these features of the factual position. In consequence, the Judge failed to address the right question, and moreover reached a conclusion on the Causation Precondition which was unsustainable, and unsupported by any evidence. The Judge simply assumed that the trial would have gone ahead. He said at [4] that "it was impossible to continue the trial" on Day 7, overlooking that it was already impossible – on the balance of probabilities – at Day 4. The jury could not have sat beyond the 10 day slot, and that was already going to be necessary. The Judge's failure to grapple with the legally relevant question – namely the counterfactuals – meant that he gave legally inadequate reasons.
69. It is legally irrelevant that Exolum's advocate (Mr Cooper KC) did not raise or make submissions on these counterfactuals. The Causation Precondition was a step which the Judge was required to take, and which he was required to address lawfully. This can be seen from Kenneally §§24, 40, where the Judge was positively invited to exercise the power by the very same claimant who successfully then challenged its exercise as involving a misdirection in law as to the nature of the power. A similar point arose in CCE (Leicester) at §37.

Discussion

70. I am unable to accept these submissions. In my judgment, the Judge's approach to the Causation Precondition entailed no material public law error. My reasons, agreeing with the submissions of Mr Puzey for HSE, are as follows:
71. The starting point is that the Causation Precondition required the Judge to ask whether HSE's incurred costs of the abandoned trial had been incurred "as a result" of the improper omission which he had found. The Judge identified the Causation Precondition. The Judge asked the question required by Regulation 3. He answered that question. His Ruling at [6] expressly recorded that question and his answer to it. It is important to remember that – as it arose on the facts and in the circumstances of the present case – the answer to the Causation Precondition necessarily involved an evaluative judgment, with a built-in latitude, for the Judge. As Mr Molyneaux rightly accepts, it was not a hard-edged question of fact (precedent fact) or law for a substitutionary supervisory review.
72. Mr Molyneaux's first 'counterfactual' concerns whether, had Exolum raised the OSA Issue in good time before the trial, a longer trial slot would have been identified. In my judgment, the answer to that is and was clear and the Judge plainly had it well in mind. The Judge had specifically addressed the length of trial slot which was necessary with the OSA Issue having been raised. It was 3 weeks (15 days) instead of 2 weeks (10 days). This assessment had been formed prior to the Ruling. It was not a product of what had happened up to Day 4. It was not a product of lost time during a trial. It was the straightforward product of the implications for the trial slot of the raising of the OSA Issue. Indeed, it was in the light of the new OSA Issue and its imponderables that the Judge was speaking of a further six to eight days. The chosen 3 week slot for the new trial was premised on appropriate steps being taken, including the position of the MOD being elicited. The Judge had considered and endorsed the view that 3 weeks was the required hearing slot for a properly prepared trial in which the OSA Issue was being raised. He had done so on Day 7 at 14:33. That was the afternoon immediately before hearing the Regulation 3 arguments and delivering the Ruling. It is not difficult to see why Mr Cooper KC did not advance this counterfactual as a basis for the refusal of the Regulation 3 costs order. It would have received short shrift. Mr Cooper KC would have been saying: 'if we had raised the OSA Issue promptly, after that two day hearing earlier this year when you dismissed our abuse of process application, it is more likely than not that you and I would have identified a 2 week (10 day) trial slot as sufficient'. The Judge's answer inevitably would have been: 'no, we would have identified a longer slot, remembering the 3 week (15 day) trial slot that you and I identified yesterday as a result of the OSA Issue being added into the rescheduled trial'. In my judgment, that is the end of the case.
73. But nor in my judgment, in any event, would Mr Molyneaux's second 'counterfactual' assist him. The Judge – and Mr Cooper KC – had their exchange on Day 4. The transcript records that, for his part, Mr Cooper KC thought it more likely that the Judge would be sending the jury out on the morning of Day 10. But the transcript also records that this position involved a pessimistic assumption being made about two contingencies. One was how long the Judge took to rule on the half-time submission. The other was how long the Judge was going to take for the summing up. On both of these, Mr Cooper KC was adopting 'worst case scenario' assumptions. And these cannot be attributed to the Judge. Each of the pessimistic assumptions concerned aspects of the trial which were in the Judge's control, on which the Judge was far better placed to have a view. The Judge's actions and words are, moreover, instructive. As to actions, the Judge decided he did not need to raise the position with the jury. As to words, the Judge said that this was because it would involve worrying them "unnecessarily". He also said as to going into a third week that "there may be an issue". His words were "unnecessarily", and "may". Again, it is not difficult to see why Mr Cooper KC did not raise this counterfactual as a basis for the refusal of the Regulation 3 costs order. It too would have received short shrift. Mr Cooper KC would have referred the Judge to their exchange and would have been saying: 'based on my prognosis of what was most likely, and what we later discovered about jury availability, this trial was on balance going to be abandoned in any event from the morning of Day 4'. The Judge's answer would inevitably have been: 'no it wasn't; there was a potential problem; but there were several actions that I could and would have taken to keep the trial effective'.
74. All of this was very fresh in the mind of the Judge, who had first-hand recent experience of these exchanges and who knew perfectly well what the position was. This is not a situation where there was any question of failing to enquire into or grapple with some relevant point. Rather, it is a case where the relevant matters involved the direct knowledge and recent experience of the decision-maker themselves.
75. The final point is about reasons. Mr Molyneaux is right to say that the Judge's reasons do not address the position as it was on Day 4; nor his reference to 6-8 days on Day 7. The Judge did not in his Ruling go into the point about why 10 days would, on balance, have been identified as inadequate had the OSA Issue been raised promptly before trial. He did not go into the point about why the position and discussion at Day 4 did not show that the trial was likely, on balance, to have been abandoned in any event. The legal standard applicable in public law, to assess whether legally adequate reasons have been given by decision-maker, is a practical one. It recognises that decisions are being given to an informed audience. It identifies as a key part of the standard of legally adequate reasons the need to address the principal controversial issues. None of what is now relied on was a principal controversial issue.
76. In these circumstances, and for these reasons, I reject the submissions that there was any material public law error on the part of the Judge in his approach to the Causation Precondition. The question of a parasitic human rights violation – an agreed consequence between the parties had I answered the issue in Exolum's favour – does not arise and I say no more about it.

Remedy

77. If I had accepted that Exolum could point to a material public law error vitiating the Judge's decision, a further agreed issue would have arisen as to what – if any – remedy the Court should then grant. Mr Puzey questioned the utility of the legal challenge, given that Exolum was convicted at the subsequent retrial and ordered to pay the costs of both trials. Exolum's answer is that if its pending appeal against conviction at the retrial succeeds, that costs order will be overturned so that this issue has a contingent utility. Mr Puzey then raises a number of points about Exolum's general approach to the criminal proceedings. But these would not have been a basis in my judgment to refuse a remedy, had I found a material public law error. Especially given the agreed position as to a consequential human rights breach. That leaves Mr Cooper KC's failure to raise the 'counterfactual' arguments now relied on. That could be a basis for refusing a remedy as a matter of discretion, in line with the analysis and outcome in TM Eye. However, this point already features in my analysis as to adequacy of reasons. I find it artificial to posit the claim succeeding, but this same point then featuring adversely as to discretion and remedy. I am not able to say that – had I found a material public law error – I would still have refused a remedy as a matter of discretion.

Outcome and Consequentials

78. Having circulated this judgment as a confidential draft, I am able to deal here with the Order and any consequential matter which arises. I will order that the claim is dismissed. I will order that Exolum pay the HSE's costs assessed at £17,557.50, from which is deducted £400. That deduction constitutes the assessed costs of Exolum's response to the HSE's application for an extension of time for its Detailed Grounds. Those costs were ordered earlier in the proceedings, when permission for the extension of time was granted. These assessed sums are all agreed, subject to one contested point. Exolum invites a reduction, at a suggested level of 30%, as a marker of disapproval for the lateness of the HSE's Detailed Grounds. The invitation is not improperly made. But acceding to it is not, in my judgment, appropriate or proportionate in all the circumstances of the present case. A candid explanation was given. The extension was obtained. An appropriate costs order was made. There was no prejudice. No 'marker' is needed and, given that, the reduction would be an unjustified windfall for Exolum.