England and Wales High Court (Administrative Court) Decisions

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
DIVISIONAL COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		07/07/2017

Sir Brian Leveson P:

1. This is the judgment of the court to which we have both contributed.
2. For many years, notwithstanding the provisions of s. 3(1) and 7A(2) of the Criminal Procedure and Investigations Act 1996 ("the 1996 Act"), disclosure of material unused by the prosecution in criminal cases that might undermine the prosecution case or support the case advanced by the defence has continued to give rise to serious issues. On the one hand, it is beyond argument that miscarriages of justice have occurred where such material has been withheld. On the other hand, attempts by the defence to seek unnecessary or inappropriate disclosure run the risk of undermining the trial process; as Lord Bingham made clear in R v H & C [2004] 2AC134, [2004] UKHL 3 at [35]:

"If material does not weaken the prosecution case or strengthen that of the defendant, there is no requirement to disclose it. For this purpose the parties respective cases should not be restrictively analysed. But they must be carefully analysed, to ascertain the specific facts on which the charges are resisted. The trial process is not well served if the defence are permitted to make general and unspecified allegations and then seek far-reaching disclosure in the hope that material may turn up to make them good. Neutral material or material damaging to the defendant need not be disclosed and should not be brought to the attention of the court."

3. These cases concern two prosecutions brought against motorists in unconnected circumstances for driving a motor vehicle on a road or other public place after consuming so much alcohol that the proportion if it in their respective breath exceeded the prescribed limit, contrary to s. 5 of the Road Traffic Act 1988. The same defence solicitors have appeared and, in both cases, defence statements have been served which deny the consumption of sufficient alcohol to give rise to a positive reading and challenge the reliability of the Lion Intoxilyzer device used in the procedure. Pursuant to these statements, applications have been made under s. 8 of the 1996 Act for comprehensive documentation concerning the relevant device, relying on expert evidence to the effect that there must have been some defect in the device: the evidence proceeds on the unstated premise that what is said by each of the motorists as to their alcohol consumption is accurate.
4. In the event, in each case, the District Judge acceded to defence applications for disclosure of this material on the basis that once an issue as to reliability of the device had been raised and documentation requested, the prosecutor was under an obligation to consider reliability, inspect the material and, disclose it. In one case (that of Joshua Whyte), after considerable difficulty, the Crown Prosecution Service ("CPS") complied with the terms of the order. [CURRENTLY REDACTED]

Jurisdiction

5. Before dealing with the merits of these applications, a preliminary issue arises in relation to jurisdiction. In particular, Jeremy Benson Q.C. for the interested parties argues that the decisions that are in issue are interlocutory case management decisions which this court does not have jurisdiction to hear. If the material disclosed does undermine the prosecution case, the interested party would properly be acquitted; if he is convicted, the remedy for the prosecution is compensation in costs.
6. In support of that proposition, Mr Benson relies on R v Rochford Justices ex parte Buck (1978) 68 Cr App R 114. The facts of that case concerned the offence of permitting the unloading of duty free goods from aircraft. The defendant faced five charges but, following objection, one charge was heard first. During the trial, an application by the prosecution to adduce the facts forming the other charges as similar fact evidence was refused and, when the prosecution sought to review that decision, the trial was adjourned. This court (Lord Widgery CJ, Croom-Johnson and Stocker JJ) held that the prosecution should have continued with the case to its end and then, if necessary, proceeded to the Divisional Court by way of case stated. The headnote reads that it was held that there was no jurisdiction in the Divisional Court to interfere with the justices' decision, which had not been reached by termination of the proceedings below.
7. Citing Carden (1879) 5 QBD 1, in which Lord Cockburn CJ observed that the court had "no authority … to control the magistrate in the conduct of the case or to prescribe to him the evidence which he shall receive or reject", Lord Widgery went on to deal with the type of proceedings being determined and observed (at 118):

"The obligation of this Court is to keep out of the way until the magistrate has finished his determination seems to me to be a principle properly to be applied both to summary trial and to committal proceedings."

8. It is worth underlining that both Lord Cockburn and Lord Widgery were dealing with appeals during the course of the hearing of the relevant trial (thereby causing an adjournment for the decision to be challenged.) We consider that these decisions can be explained (and justified) on that basis. The same is so for the subsequent decision which relied on Buck. In R (Hoar-Stevens) v Richmond Magistrates Court [2003] EWHC 2660 (Admin), the defence wished to explore the reliability of the breath device and had sought access to records in relation to it. The prosecution had not fully complied with orders relating to disclosure and, at the conclusion of the prosecution case, a submission was made that the prosecution should be stayed as an abuse of process. The District Judge considered she was not bound by the earlier disclosure ruling and refused the application, reserving the question whether to exclude the material pursuant to s. 78 of the Police and Criminal Evidence Act 1984 to be addressed during closing submissions. This court considered itself bound by Buck, Kennedy LJ observing (at [18]):

"The proper course is to proceed to the end of the trial in the lower court and then to test the matter, almost certainly by way of case stated."

9. John McGuinness Q.C. for the D.P.P. accepts what he describes as this general principle but challenges the proposition that it is a question of jurisdiction. Thus, in R(Watson) v Dartford Magistrates' Court [2005] EWHC 905 (Admin), following a refusal by one bench to adjourn a trial, a further application without a change of circumstances was granted. On appeal to this court (Sedley LJ and Mitting J), it was not accepted that there was a bar to intervention. Referring to Buck, Mitting J put the matter (at [7]) in this way:

"I accept that this is the normal rule, but in this, as in some other cases, the prosecution would no doubt say at that conclusion of a trial resulting in a conviction that it was too late for the claimant to complain about an adjournment that should not have been granted before. In a case such as this, where the issue is straightforward and the principle clear, I do not see that there is any fetter on this court intervening."

10. Similarly, in CPS v Sedgemoor Justices [2007] EWHC 1803 Admin, the court was concerned with an excess alcohol case in which the magistrates ruled that the evidence of a toxicologist was inadmissible as she was not an 'authorised analyst'. Dealing with the application for judicial review, the court (Hughes LJ and Treacy J) referred to Buck, Hoar-Stevens and Watson. Hughes LJ (as he then was) went on (at [11]):

"It needs to be said, it seems to me, that the general rule as set out in Buck is important. It is necessary in nearly every case to wait until the end result of the trial in the Magistrates' Court is known before anybody can say whether there is a source for complaint or not. That was the position in Hoar-Stevens because other decisions - for example, that under Section 78 of the Police and Criminal Evidence Act - might yet remove any reason for complaint at all. Quite apart from that, it is necessary in nearly every case to wait until the end of the case so that the magistrates can find the facts and they can properly be known for the purposes of decision here. It is for that reason that it is the procedure by way of case stated which is, in nearly every case, the correct manner in which to challenge a decision in the Magistrates' Court. Otherwise the facts may be the subject of dispute in this court."

11. Mr Benson points out that Hughes LJ went on to say (at [12]) that the application for judicial review ought not to have been brought and that the prosecution should have behaved as in Buck. That is so, but the court must have accepted that there was no absolute jurisdictional bar because (albeit with some hesitation) it did determine the matter and did so in the prosecution appellant's favour. This approach followed what Hughes LJ had said in R (Singh) v Stratford Magistrates Court [2007] EWHC 1582 (Admin) dealing with a substantive application that the magistrate be required to conduct a trial which it was contended he was declining to do. Citing Watson, he observed (at [8]) that the court had sometimes been persuaded to consider a case which was interlocutory where there was good reason for doing so but that this should not be taken as any encouragement, for such applications were likely to be dismissed "out of hand". He went on to say that had the point been spotted earlier, it would likely to have been regarded as premature but, critically, it was, in fact, determined.
12. The first answer to Mr Benson's submission is that the decisions in the present cases (following an interlocutory ruling and well in advance of the trial) were not made during the trial itself and, thus, strictly the Buck principle can be distinguished: it is not difficult to see why the court will do all that it can not to interrupt a trial then proceeding to a conclusion. In any event, we doubt that the proposition that there is a true jurisdictional bar (meaning that the court had no right to consider the issue at all) can be justified: see, for example, the observations of Sedley LJ in Essen v DPP [2005] EWHC 1077 (Admin) (at [38]) which suggested that the Buck group of decisions could usefully be revisited on the basis that a fixed rule that any challenge must abide a final outcome is capable of working injustice.
13. There are obvious reasons why the more recent cases were, in fact, determined on their merits (contrary to what would be a true jurisdictional bar). Including words such as 'generally' and observations such as 'in nearly every case' underline what is an entirely pragmatic response to the modern approach to case management and the conduct of hearings in the magistrates' court. In our judgment, however sensible the general rule is (almost inevitably so if a challenge is mounted during the course of the trial), in appropriate and exceptional cases, a mechanism that permits a challenge is entirely consistent with the overriding objective identified in the Criminal Procedure Rules. Accordingly we are satisfied that we have jurisdiction to hear these cases.
14. In the circumstances, it is appropriate to restate the approach in this way. First, it is difficult to visualise circumstances in which it would be appropriate to adjourn a trial simply for the purpose of challenging an interlocutory ruling made during the course of that trial. Such a challenge should be pursued at its conclusion. Second, a challenge to an interlocutory order or decision should not lightly be made but may, exceptionally, be justified where the challenge raises issues likely to have general or wider application and is not dependent on the ultimate result and there is no other means by which the order or decision can be challenged.
15. Notwithstanding Mr Benson's submissions to the contrary, the type of challenge in this case is a good example of the circumstances in which the court should be prepared to exercise its jurisdiction to intervene. The proper approach to applications of this nature (whether they should be acceded to or rejected) requires principled resolution not least to ensure that practice is consistent across the country. If, as a matter of principle, an order of disclosure is justified, that is one thing. If, however, it is not justified, that is different. To argue that an inappropriate order in relation to disclosure can be justified because it might be compensated in costs runs counter to the necessary principled resolution of issues in a way that prevents repetitive arguments in different courts, with none providing a binding solution to the problem.
16. That point is amply illustrated by the evidence in this case. Ms Cronshaw, Senior Prosecutor in the CPS who deals with most of the Greater Manchester excess alcohol cases, provided a witness statement explaining the concerns of the CPS about a pattern of disclosure requests of this sort now being made routinely in excess alcohol cases. Arguments about disclosure had taken up significant court time and had caused a number of adjournments even after the trial had started. In one case, after several adjournments, and with the disclosure refused, there was a conviction. In another disclosure was ordered but not complied with, leading to the exclusion of the breath test evidence and hence to acquittal. The sort of request seen in the case of Mr Whyte was now routine. The CPS received such requests now once every three weeks, regardless of previous responses or case management directions. The letters of request were largely standardised and almost always sought the metrological log, subject tests, calibration certificates and engineers' reports. There were serious practical implications for the resources of the CPS and police. The disclosure arguments took up a great deal of court time, with increased costs, delays, and adjournments, affecting police and civilian witnesses. It is often said, on the day of trial, that the material should have featured on the schedule of unused material, and that the CPS has not complied with its disclosure obligations.
17. The lawfulness of the disclosure orders could not sensibly be challenged by the CPS in the event of a conviction. Where there was an acquittal, the disclosed material may have played no part in the acquittal; it might not even have been admitted in evidence. The CPS is bound to comply with court orders, even if it intends to challenge them later if possible; it should not wait until it faces the consequences of non-compliance, for example in the refusal to admit the printout in evidence, and then seek to remedy the consequences of non-compliance by challenge to the order it failed to comply with. A costs order if made, and paid, will not establish any general principle in relation to the making of a disclosure order; and will not cover the disruptive effect of delays on the trial process.
18. In Cracknell v Willis [1988] AC 450, Lord Goff was correct (at 471H) not to share the optimistic view that motorists would desist from seeking to persuade magistrates to reject evidence from printouts as unreliable on the ground that they have drunk so little that the reading cannot be right and (at 472C) to anticipate the need to keep under very careful review the situation arising from that decision. How the courts should address applications for disclosure made pursuant to s. 8 of the 1996 Act is one such opportunity not, we emphasise, necessarily to limit such applications but rather to ensure that they are considered and approached in conformity with principle and a proper application of the law.
19. In the circumstances, we reject the challenge to our jurisdiction to hear these cases and turn to consider the judicial review applications on their merits. Ouseley J ordered both to be dealt with as rolled-up hearings. We give permission.
20. The remaining issues require, first, an analysis of the provisions of the 1996 Act to determine whether the disclosure ordered falls within its compass and, if so, which provision. The second issue is whether, even if the material is capable of falling within the 1996 Act, the District Judge could rationally conclude on the sort of evidence he had that it was reasonably capable of undermining the prosecution case or assisting the defence case.

The facts: Mr Whyte

21. The prosecution case is that the police stopped Mr Whyte when he was driving his car in the early hours of Saturday 3 September 2016. He appeared to demonstrate slow and slurred speech; there was the smell of alcohol in the car, and on his breath. He was unsteady on his feet. He gave a sample of breath for a preliminary roadside test, which yielded a positive reading of 61 micrograms (µ), in 100 millilitres (ml) of breath. At Wigan Police station, the investigative evidential breath procedure was carried out between 04.25 and 04.45 using a Lion Intoxilyzer 6000 UK device, type approved for this purpose. The print out showed no error; the two readings were 64 and 61 µ of alcohol to 100 ml breath. The lower one still exceeded the prescribed limit, which is 35, although the individual is not prosecuted unless the alcohol level reaches 40. Mr Whyte was charged with an offence contrary to s5 RTA 1988.
22. On 27 January 2017, Mr Whyte served his defence statement pursuant to s6 CPIA 1996. He said that he did not drive whilst he was over the prescribed alcohol limit and it asserted that "He takes issue with the reliability of the breath alcohol readings at the roadside and in the police station…". It went on that he had consumed insufficient alcohol to give rise to those readings or to any exceedance of the prescribed limit. He referred to two expert reports, from a Dr Mundy and his successor Miss Dale, to the effect that the amount he said that he had drunk between 19.30 and 01.55 on 3 September 2016, for a man of his height and weight would have yielded a zero reading. The experts "had raised concerns over the reliability of the evidential specimen and noted that this is incompatible with the roadside instrument". The experts requested disclosure of 7 items which were then set out.
23. Dr Mundy, an independent forensic alcohol consultant, who served in the Forensic Science service until retirement in 1999, and who has appeared in a number of cases such as this, is a specialist in the field of alcohol detection and measurement in biological specimens such as blood. He said that the roadside test was incompatible with the evidential tests because the former, though earlier, showed a lower percentage of alcohol in the breath than the later tests. He said in his 12 December 2016 report that part of the disclosure relating to the evidential breath tests was required "to show that the results…are reliable." He continued "If it becomes necessary in order to verify that the Intoxilyzer 6000UK was in full working order and reliable" the rest of the disclosure identified in the defence statement would be required. The focus, though the request may just have been badly worded, was on the contrast between the results of the roadside test and the evidential specimen tests, which meant that either or both were unreliable; but he also did refer to the zero presence of alcohol in the breath on what Mr Whyte said of his drinking.
24. Miss Dale, who had also worked in the Forensic Science Service until 2001, and had then also become an independent forensic alcohol consultant, agreed with Dr Mundy's conclusions in her report of 31 December 2016, both as regards the calculation of alcohol in the breath on the basis of what Mr Whyte said, and on the contrast between the roadside and evidential tests: "This suggests that either one, or possibly even both, of them is incorrect which, as a consequence, raises a question mark over their reliability."
25. The CPS, which did not receive the two reports with the defence statement, replied to the disclosure request on 30 January 2017, saying that the statement had not complied with s6A of the 1996 Act. The CPS sought a compliant defence statement. It received Miss Dale's report shortly afterwards but not Dr Mundy's. Nonetheless it replied on 7 February 2017 saying that none of the documents requested were unused material, being neither obtained or inspected in the course of the criminal investigation, nor relevant since they were neither capable of undermining the prosecution case or assisting the defence. The statement had not explained why each document was relevant or why each could assist. The key document for the purposes of the evidential specimen tests was the print out from the machine which had already been supplied.
26. The issue of disclosure was raised at a hearing on 9 February, which caused the trial to be adjourned. On 23 February 2017, Mr Whyte served an application for disclosure under s8 CPIA. This specified 13 items:

"1. A complete MG6C which particularises all material that may be relevant to the investigation

2. Details of the previous and post checks and calibrations used for the "roadside breath test" prior to the 3^rd September 2016

3. Details of the results of the calibration checks of the "roadside breath test" prior to the 3^rd September 2016

4. A copy of the printout from the download which may contain information for that 35 day period around the 3^rd September 2016 on the "roadside breath test".

5. Disclosure of the download of the memory showing the subject test; and the tests 6 months before and 6 months after 3^rd September 2016 on the Intoxilyzer.

6. The metrological log – and details of the 299 tests on the download memory of the Intoxilyzer.

7. Relevant certificates of calibration from the Intoxilyzer and gas cylinder before and after Mr Whyte's breath procedure.

8. Any engineers reports from seven months either side of Mr Whyte's breath test from the Intoxilyzer.

9. Certificates of the check gas used in the test at the police station.

10. Copies of service certificates of the Intoxilyzer 6000UK number A0358 before and after 3^rd September 2016 plus the engineer reports for the same period.

11. A report from the roadside device if it exists.

12. All relevant pocket notebooks not previously disclosed.

13. Any pre-release breath test printouts or confirmation they do not exist."

The application explained in general terms that the prosecutor was thought to have that material because the "metrological log book" had to be kept with the device, and would record the results of service and repair. As the device had to be calibrated, the calibration certificates had to exist in order for it to retain its type approval.

27. The material was said to be capable of undermining the prosecution or assisting the defence because the two reports said that Mr Whyte's breath may have truly been below the prescribed limit, and they therefore raised an "inferential challenge to the reliability of the device and the above material is necessary to allow our expert to form a definitive opinion in relation to the same". The roadside reading was incompatible with the evidential readings.
28. The CPS rejected this application. The essence of its response was that the items sought were not unused material, as they were not obtained or inspected in the course of the investigation into Mr Whyte's alleged offence, and there was no specific explanation of how each item was said to be unused material, nor had any basis been shown for each item to come within s8(3) or (4) of the 1996 Act.
29. At the hearing in the Manchester and Salford Magistrates' Court, after argument from both sides, DJ Hadfield ordered disclosure of four items: (1) the metrological log for the device in question, that is the record of servicing and repair normally kept in the custody suite; (2) the calibration certificates before and after the date of use in question 3 September 2016, and also normally kept with the device in the custody suite; (3) the results of the previous 299 tests in which the device was used and which were retained in its memory; and (4) the service certificates and engineer logs before and after 3 September 2016. These are items 6, 7 and 10 on the list requested. It appears that these duplicate items 5 and 8 as well. DJ Hadfield ordered disclosure because he judged that the defence had raised an issue about the reliability and functioning of the evidential breath device. He refused to order disclosure in relation to the roadside testing device as it could not assist the defence in any way.
30. The memory of the machine can only be downloaded either by Lion Laboratories or by someone with the necessary technical expertise, of which there is only one in the Greater Manchester Police. He undertook that task. The service certificates and engineer logs were obtained by request to Lion Laboratories, which they complied with. The other items were obtained from the custody suite. Accordingly, the required disclosure was provided, and is before us. Ms Cronshaw said that in Mr Whyte's case alone significant hours and effort had to be devoted to obtaining this material.
31. The trial date was adjourned from 26 April to a mention on 3 May 2017, because of witness availability. On 2 May, these judicial review proceedings were lodged.
32. [CURRENTLY REDACTED]
33. [CURRENTLY REDACTED].
34. [CURRENTLY REDACTED]
35. [CURRENTLY REDACTED]

The further evidence

36. Ms Cronshaw explained that the CPS did not regard the disclosure as pursuing a reasonable line of enquiry, drawing upon a statement dated 11 April 2017 from Dr Williams, which we considered de bene esse. He is an expert in the field of body alcohol, but the relevance of his evidence derived from his experience of the Lion Intoxilyzer 6000 UK; he was responsible for its original design specification, co-ordinating its design and engineering processes and developing the device to UK requirements. He explained the operation of the device in terms which were not disputed before us, and are likely to be familiar to District Judges:

"F.1 Purging and Blank Checks

Before the Intoxilyzer can analyse a person's breath or carry out a calibration check it is a Home Office requirement that is it clear of alcohol from any previous usage, and that it is therefore reading 0µg/100ml as a baseline. To ensure that this is the case the instrument purges itself by drawing in room air through the breath tube and mouthpiece: it then checks that the background reading is 0.6µg/100ml or lower. If during this purging process the 6000 is unable to score a reading of 0.6 or lower then the message Ambient Fail is shown on the display and printout, and the test procedure automatically terminates at that point. This same purging and air blank checking routine of each specimen of breath or standard alcohol vapour is quite unaffected by the alcohol in whatever specimen the instrument may previously have received and analysed.

F.2 Calibration Checks

This calibration check procedure is carried out before AND after the subject provides his or her two breath specimens, to ensure that these are each analysed accurately for their alcohol content. It is conducted using an alcohol vapour/air mixture of known concentration, nominally 35µ/100ml, held in a cylinder that is attached to the side of the instrument. This mixture is passed into the Intoxilyzer instrument automatically, and the reading obtained must lie within the defined range 32-37µg/100ml inclusive. In fact the range is actually 32.0-37.9, but because the measured readings are always truncated before reporting, they appear as 32-37µg/100ml on the display and printout.

Provided, therefore, the two calibration checks shown on any printout [under Simulator Check 1 and Simulator Check 2] are within the range 32-37µg/100ml inclusive, this is proof that the said instrument was analysing breath alcohol to the required limits of accuracy [as defined by the Home Office] at the relevant time."

Later he added:

"I reiterate … – the lion intoxilyzer® 6000UK printout relating to WHYTE's evidential breath analysis [Test Number: JJ/10668/16] shows that the two calibration check readings [Simulator Check 1 and Simulator Check 2] were 35 and 34µg/100ml [as I have stated at Section F.2 the allowable range is 32-37µg/100ml] and all four Blank readings were 0µg/100ml. There are no error messages and the two reported breath alcohol readings are within 15% of each other."

He concluded that there was "simply no direct evidence that [the device used for Mr Whyte] was unreliable when the defendant blew into it… the print out was remarkably unremarkable." He was critical of Dr Mundy's absence of reference to the print out, which showed no error, or to its significance.

37. He dealt with the reliability of the roadside test instrument: the type of device used could not work at all if it were faulty because the software would close the machine down automatically, as part of its type approval. But that is not at the heart of this case, since disclosure of its records has rightly not been ordered, and it is not an evidential source.
38. He then dealt with the Intoxilyzer 6000 UK, in these terms:

"The requirement of the Home Office for Type Approval was [and still is] that an instrument can only issue an evidential breath analysis record if, as a result of the various required checks that it goes through as a part of each breath analysis procedure [see Sections F1 and F2], it is shown to be operating correctly and accurately when the subject provided their two breath specimens into it. The printout stands alone: it is the apex to the pyramid. All other the documents (sic) [the building blocks of the pyramid], are enabling documents. If those processes had not been completed correctly, and on time, then there would be no printout: that is a pre-requisite to Type Approval."

39. In effect this is a device which fails safe in the sense that, if faulty, it will not work at all rather than produce readings which would be unreliable.
40. Dr Williams considered the previous test results in the memory of the device, saying that there was no need to see it or the service records; it would reveal nothing about the reliability of the reading in this case. The data of other tests could even be misleading in the absence of full knowledge of the background of those entries. Pro forma check lists, which are the format of the service reports and certificates, when the device is serviced and its calibration checked, were irrelevant: if the device were unchecked within the time limits or if it failed the checks, the device cannot be used to analyse breath. As Mr Whyte could blow into it, there was no unreliability, and no need for these documents. The gas certificate was only relevant to testing the device and not to its use; if the gas were at an incorrect value, the device could not actually be used to test breath.
41. Dr Mundy produced three further reports each dated 20 June 2017, which we also considered de bene esse. The first dealt with the significance of the material disclosed in the case of Mr Whyte. He said that out of the 299 other subject tests in the memory, there were 20 where one or other of the specimens had not been provided; 3 related to the second specimen. There also three failures because of the detection of an interfering substance in the second specimen on results between 55 and 58 µ to 100ml breath. He explained the possible sources of these interferences and how the device is designed to deal with them. But he concluded that, because the interferences were found on the second and not the first specimen, when he would have expected the level to have reduced, and the range was very small, the results were "very probably not reflective of an accurately operating instrument." This device therefore "has a problem" It was detecting interfering substances when there were none. The results were extremely unusual.
42. His second dealt with the report of Dr Williams. He raised a variety of issues including the inability of the print out to record full data of blows of breath below 12 litres a minute, based on experiments he had carried out and notified to the Home Office in 2015. In addition to the comments he had made about the logs disclosed in the case of Mr Whyte, he referred to a small number of cases in which the logs had been disclosed and instanced a case where he said that the calibration check standards in a few cases fell "explicably", (we think that he meant "inexplicably") below the acceptable alcohol level. He also said that there had been a few occasions when an out of date gas cylinder had been used, "which could raise various issues e.g. approval and reliability."
43. His third report expressed his view that the only way the performance of an instrument at the time of test could be scrutinised is by "viewing the documents associated with that instrument." The metrological log should show that the six monthly services had been carried out, and whether any faults had been reported. The subject test log could be used to give a measure of the precision of the instrument, whether it was performing normally, the performance of a specific operator could be monitored, and the use of calibration gas outside its expiry date could be checked.
44. We have decided to admit these three reports along with that of Dr Williams.

The first issue: The statutory and Code provisions

45. Section 3 of the 1996 Act contains the prosecutor's initial disclosure duty of "prosecution material"…. "which might reasonably be considered capable of undermining the case for the prosecution against the accused or of assisting the case for the accused". "Prosecution material" is defined in subsection (2) in the same way as it is in s8(3) (a) and (b), and is set out below.
46. The continuing duty of disclosure in s7A, which includes review of disclosure after a defence statement, still relates to prosecution material as defined in s3, with the same relevance test.
47. Section 8, the provision invoked before the District Judge, enables the accused who has given a defence statement, to apply to the court for an order requiring the prosecutor to disclose prosecution material to him, if he "has at any time reasonable cause to believe that there is prosecution material which is required by section 7A to be disclosed to him". Section 8 goes on to provide:

"(3) For the purposes of this section prosecution material is material—

(a) which is in the prosecutor's possession and came into his possession in connection with the case for the prosecution against the accused,

(b) which, in pursuance of a code operative under Part II, he has inspected in connection with the case for the prosecution against the accused, or

(c) which falls within subsection (4).

(4) Material falls within this subsection if in pursuance of a code operative under Part II the prosecutor must, if he asks for the material, be given a copy of it or be allowed to inspect it in connection with the case for the prosecution against the accused."

48. It is not enough for the purposes of s8(3) (a) or (b) that the material is in the prosecutor's possession or has been inspected by him. Subsection (4) is an addition to the definition of "prosecution material" found in s3.
49. For these purposes, Mr McGuinness accepted that the metrological record of service to the machine, kept with it in the custody suite, the records of the 300 tests in the memory of the device, and the calibration certificates, were in the possession of the police. (It should be noted that this concession was not made in the grounds for judicial review.)
50. However, we accept his submission that none of the material was "prosecution material" within s8(3) (a) or (b) since none of it came into the prosecutor's possession in connection with the case against the accused nor had it been inspected in connection with that case either. The service certificates and engineer logs were not in the possession of the prosecutor at all, but were obtained by request of Lion Laboratories Ltd., the manufacturer of the device.
51. Section 8(4) provided the only other route for statutory disclosure. There was no obligation on Lion Laboratories to provide the service certificates and engineer logs to the prosecutor, if asked, or to make them available for inspection. The subsection focuses on the obligation to supply, not on any obligation to make a request. So those items fall outside s8(4).
52. The documents in the custody suite, including for these purposes the memory within the device, could only come within s8(4) if they were held by the Investigating Officer or by the Disclosure Officer; see the reasoning in DPP v Wood and McGillicuddy [2006] EWHC 32 (Admin), [50 and 55]. Once it is accepted, as it was here, that they are in the possession of the police, it was not suggested that they would fall outside s8(4) upon a request being made by the prosecutor pursuant to an operative code. The relevant Code is the March 2015 CPIA s23(1) Code of Practice issued by the Ministry of Justice. Paragraphs 3.5-3.6 in the section headed "General Responsibilities" require the investigator to "pursue all reasonable lines of enquiry whether these points towards or away from the suspect." If there is an obligation on the prosecutor to ask for the material under that provision, s8(4) would be satisfied. But there would still be no obligation under either s3, or under the continuing obligation in s7A, to disclose such material unless it was reasonably capable of undermining the prosecution case or assisting the defence. And, of course, if it were known that that was the position, it would not be a reasonable line of enquiry to pursue either. We therefore turn to the second issue.

Could the District Judge rationally conclude on the evidence that these materials were reasonably capable of undermining the prosecution or advancing the defence cases?

53. In our judgment, the answer to that is no, and disclosure should not have been ordered here. There was no basis upon which they could reasonably have been thought capable of undermining the prosecution or advancing the defence case.
54. First, those seeking and those making disclosure orders in excess alcohol cases must bear in mind the risks to which Lord Goff spoke, as set out above in Cracknell v Willis, above. These have been brought home recently by the decision of the Divisional Court in R (Hassani) v West London Magistrates' Court [2017] EWHC 1270 (Admin) and the appended extracts from the judgment of Senior District Judge Riddle in CPS v Cipriani. This means that there must be a proper evidential basis for concluding that the material sought is reasonably capable of undermining the prosecution or of assisting the defence, or that it represents a reasonable line of enquiry to pursue. We appreciate that DJ Hadfield did consider the extensive disclosure request because, plainly rightly, he declined to order disclosure of much of what was sought. We accept that he heard argument and asked some questions of Miss Dale. But we are satisfied that there is no evidential basis upon which the disclosure should have been ordered.
55. Second, it is not enough for one or more experts to say that the material is necessary to verify that the device was reliable in the language used in the reports of Dr Mundy and Miss Dale in support of the application for disclosure. Nor does the written application for s8 disclosure provide any evidential basis for it. It is not enough to say that the defence case is that the amount drunk would not put the defendant over the limit or anywhere near it, and therefore the machine must be unreliable. What the evidence needed to do, in order to provide a basis for such a disclosure order was to address two critical features.
56. The first requirement is the basis for contending how the device might produce a printout which, on its face, demonstrated that it was operating in proper fashion, but which could generate a very significantly false positive reading, where, on the defence case, the true reading would have been well below the prosecution limit. The second requirement is to identify how the material which was sought could assist to demonstrate how that might have happened. Those are the two issues which arise and which the expert evidence in support of disclosure should address. Unless that evidence is provided, the disclosure is irrelevant.
57. But Dr Mundy and Miss Dale simply ignored the printout, and the way in which the device is designed and operates. They provide no explanation as to how the device could have malfunctioned in the way they say it must have done and still have produced a positive reading so far from the true reading which they say it would have produced. No explanation of any sort is offered for the four blank readings and the two simulator readings all designed to demonstrate that the machine is operating correctly. Their generalised assertions that the machine could be unreliable and that its reliability needed to be verified was accompanied by no evidence that the disclosure sought could cast any light on its reliability in that way.
58. Third, this is not to say that the machine must be taken to be infallible. Cracknell v Willis permits evidence to be given that the defendant had not been drinking anything like enough to produce a positive reading, even if he cannot demonstrate how the machine might have malfunctioned. But these disclosure applications go further and are addressed to identifying how that might have happened. However, unless the disclosure application addresses the two questions which we have identified, this extensive disclosure would have to be given in every case in which a defendant alleged that his alcohol consumption had been too low to sustain a positive reading, and in effect proof of reliability would always be required and the presumption of accuracy would be displaced.
59. Fourth, the comments of Dr Mundy on the material which was disclosed in Mr Whyte's case confirms all our concerns. Dr Mundy still does not suggest how that material advances a case that the machine was unreliable or malfunctioning in such a way that the positive reading may have been given when it should have been a negative reading. He raises no more than a question over the results of a few other tests, which could not be pursued without the full details of the circumstances of those tests, and where the problem as he describes it, and if problem there was, appears to be that it may have over-calculated the interfering substance and so discounted the specimen.
60. Finally, if the DJs required disclosure of material in the hands of Lion Laboratories Ltd, a third party, the appropriate procedure should have been invoked. Orders cannot be made requiring the CPS to obtain what it has no right to obtain, on pain perhaps of the prosecution being dismissed. It is not to the point that the third party may prove quite co-operative.
61. Our judgment is confined to this particular type of breath test device. Others (the Camic Datamaster and the Intoximeter EC/IR) are also in use: these may or may not have materially different characteristics.

Conclusion

62. [CURRENTLY REDACTED]. We grant no relief, beyond our declaration (at [54] above) that the disclosure order of DJ Hadfield should not have been made, in respect of his order in Mr Whyte's case, since disclosure has taken place. It will be for the trial judge to decide on its admissibility and what weight should be attached to it. Having said that, however, it has been relevant to the issues to consider the course of proceedings, and the terms of the disclosure application and order.
63. As for costs, we make no order in relation to the appeal in the case of Mr Whyte: although he contested the proceedings and thus might be considered fortunate, he need have taken no part given that the order had been complied with. [CURRENTLY REDACTED].