A necessary evil? Polygraph interviews,SHPOs and the scope of prohibition requirements: R v David ( Owen Huw ) [2023] EWCA Crim 1561

The facts of the case

The appellant was sentenced in the Crown Court at Cardiff to two years and eight months’ imprisonment in respect of the first charge (breach of a Sexual Harm Prevention Order, s. 354(1) and (4) Sentencing Act 2020) and a consecutive term of six months’ imprisonment in respect of the second charge (making indecent photographs of a child, ss. 1(1)(a) and 6 of the Protection of Children Act 1978). A Sexual Harm Prevention Order (hereafter: SHPO) was imposed until further order. The SHPO provided in relation to Prohibition 13 that the appellant is prohibited from a list of activities, ‘save where there is good reason not to do so to comply with instruction. To comply with any instruction given by [the appellant's] offender manager/police requiring [him] to attend polygraph/integrity screening. To participate in polygraph/integrity screening examinations as instructed and comply with any instruction given to [him] during a polygraph/integrity screening session by the person conducting the assessment’ (at [10]).

The sole ground of appeal for which leave was granted was that Prohibition 13 in the SHPO was disproportionate and that the Learned Judge was wrong in principle to impose such a term within the SHPO.

The prosecution supported the imposition of Prohibition 13. They argued that considering the very high risk of serious harm posed by the appellant, Prohibition 13 was not disproportionate or oppressive. What is more, in the Crown’s Respondent’s Notice it was accepted that the term could have been drafted better so that it made sense, and with changes to its ambit. They also agreed that the Learned Judge did not expressly state that he was covering the conditions to be found in ss. 343 and 347A of the Sentencing Act 2020. They submitted that the requirements of ss. 343 and 347A were met on the evidence before the court. It was also submitted that the inclusion of the words ‘save where there is good reason not to do so, to comply with the instruction’ (characterised as the ‘good reason defence’) prevented the requirement operating in a disproportionate or oppressive manner.

The counsel for the defence submitted that Prohibition 13 was not necessary and in any event it was disproportionate. It was pointed out by the defence counsel that Prohibition 13 places a positive obligation upon the appellant to participate in ‘polygraph/integrity’ screening at any time upon a request by the police; that the remit of any topics to be covered during that testing is not defined; that there is no test that must be met before the appellant could be asked to attend to participate, and his failure to participate, to ‘co-operate’ or to answer questions will amount to a breach of that order; that the broad reference to ‘your offender manager/police’ is too wide and imprecise to fulfil the requirements of s. 347A of the Sentencing Act 2020.

The Court of Appeal did not consider the points made by the appellant in relation to Articles 6 and 8 to be of any merit. For one, Article 6, Bryan J noted, is not engaged until either the criminal or civil rights of an individual are being determined, and that is not the consequence of such a positive requirement, which is being used as a safeguarding tool. For another, in the context of any prohibition imposed, any interference will be in accordance with law and justified as permitted by Article 8. However, Bryan J added, the Crown Court judge did not set out in his sentencing remarks how and why he considered Prohibition 13 to be justified, or satisfy himself as to whether the parameters of the order were sufficiently clearly delineated. In fact, all the Learned Judge said about the SHPO that he intended to make was: ‘I make a Sexual Harm Prevention Order in the amended terms sought until further order’ (at [14]).

The Court of Appeal considered thus that there is substance in the grounds of appeal so far as they relate to the imposition of Prohibition 13 and the terms thereof. It is incumbent upon a sentencing judge, Bryan J stressed, to address expressly the requirements of s. 343 of the Sentencing Act 2020, and also to give reasons as to why such an order is necessary for the specified purposes and why it is proportionate and sufficiently clear. Such reasons need not be long, but they should address such matters (at [22]). Had the judge performed such an exercise, he would have recognised that Prohibition 13, as sought and ordered, did not comply with the requirements of s. 347A of the Sentencing Act 2020 and was drafted in terms which were far too wide and vague, with the result that it was both disproportionate and oppressive. Contrary to s. 347A, Prohibition 13 failed to specify a person who was to be responsible for supervising compliance with the requirement, referring instead to ‘your offender manager/police’. This is far too vague and, on its face, would apply to any offender manager or any police officer anywhere in England and Wales and whether or not they had any responsibility in relation to the appellant. Even more fundamentally, it purports to require the appellant for the rest of his life to comply with a requirement of any police officer or any offender manager to take a polygraph test for any reason whatsoever, or indeed for no reason at all. The remit of any topics to be considered during that testing are not defined, as they should have been (at [23–24]).

The inclusion of the ‘good reasons defence’ did not save Prohibition 13. It would be difficult to apply in any given set of circumstances. It imposed a burden on the appellant to assess whether there is or is not a ‘good reason’. In some circumstances it would be difficult for him or his advisers to predict whether there was a good reason. The burden for the appellant would then be that if he gets it wrong (because a court finds that there was no good reason), he might be in breach of Prohibition 13. Far from saving Prohibition 13, the Court of Appeal considered that the ‘good reasons’ provision in fact creates more problems than it attempted to avoid, however well-intentioned it was.

Prohibition 13 is disproportionate and should be deleted from the SHPO (at [28]).

The appeal against sentence is allowed (at [29]).

Comment

One of the most interesting developments in the penal system of England and Wales is the proliferation of pseudo-scientific methods, blurring the lines between informed decision-making processes and risk assessment instruments (RAIs) on the one hand and fact-finding dystopia/science fiction on the other hand. More specifically, the notorious polygraph is enjoying an unmistakable resurgence of usage. In England and Wales polygraph interviews are currently in use by His Majesty's Prison and Probation Service (hereafter: HMPPS) to monitor among other things sex offenders on parole and manage their level of compliance. ¹ Parallel to that, the UK Parliament ² conferred new powers to judges who can now include a positive requirement in an SHPO provided that the latter is ‘necessary’ to protect the public especially children and vulnerable adults from sexual harm. ³ The combination of those two elements lies at the heart of this case.

Accordance with the law

The Court of Appeal rejected the appellant's claim that his Articles 6 and 8 rights had been violated by the imposition of the polygraph requirement. As Bryan J noted, ‘any interference will be in accordance with law and justified as permitted by Article 8’ (at [20]), equating thus the SA 2020 with the ‘law’. This, however, misrepresents well established domestic and human rights law. The Court of Appeal consistently accepts that any restriction of liberties might be ‘lawful’ in view of domestic law, but still be deemed arbitrary in view of the Convention. ⁴ In other words, compliance with the ‘law’ cannot be reduced to a single piece of (domestic) legislature. The general principles salient in the ‘in accordance with the law’ standard were recently summarised by the Divisional Court ⁵ and approved on appeal in Bridges v CC of SW. ⁶ Briefly, the measure in question must:

have ‘some basis in domestic law’; and

be ‘compatible with the rule of law’, which means that it should comply with the twin requirements of ‘accessibility’ and ‘foreseeability’, ⁷

for a legal rule needs to be both accessible and foreseeable to prevent arbitrariness in the application of the law. Any exercise of power that is unrestrained by law is not ‘in accordance with the law’. The respective court must thus carry out a close examination of the procedural safeguards embedded in the system to prevent arbitrary encroachment on individual rights including Article 8. ⁸ As the Strasbourg Court notes, domestic law must be sufficiently clear to give citizens an adequate indication as to the circumstances in, and the conditions on which public authorities are empowered to resort to any such measures. ⁹ As Bryan J himself stressed, the SHPO requirement (Prohibition 13) was ‘far too vague and on its face would apply to any offender manager or any police officer anywhere in England and Wales’ (at [24]). Neither the duration of the requirement nor the content of any topics to be considered were defined. The Court accepted that Prohibition 13 was oppressive in view of its vagueness, but failed in my opinion to recognise that it violates Art 8 too.

Signal or noise?

Now, the first question is: Was the vagueness of the requirement which the Court of Appeal deemed oppressive, within the boundaries of the law? Note that the specified period for the Prohibition can lawfully be an indefinite period (s. 347(2) SA 2020, subject to s. 348(1) of the same Act). What is more, an SHPO requirement must indeed specify a ‘person responsible for supervising compliance with the requirement’. That person, however, may be ‘an individual or an organisation’, s. 347A SA 2020. Notwithstanding, Bryan J justifiably, in my opinion, criticised the fact that the Crown Court judge did not provide reasons for the decision to impose a requirement (see s. 347A(2) SA 2020).

Second: Is the aforementioned vagueness an aberration or normality, signal or noise? To answer this question, we need to take a thorough look at the polygraph. As every blue-ribbon committee on the validity of the polygraph has concluded, the ‘stress response’ to be measured by the polygraph can be triggered by a host of factors. There is simply no unique physiological indicator that reflects a single underlying process, let alone deception (see e.g. British Psychological Society, 2004; National Research Council, 2003). It is precisely for this reason that polygraphic data are not admissible in criminal courts ¹⁰ around the world, and that even para. 4.70 of the Polygraph Instructions states that: ¹¹

[a]n examination result alone cannot be used to justify recall to custody. Regardless of the outcome of the test—Significant Response, No Significant Response or Inconclusive—this alone cannot be used to recall an individual to custody.

The polygraph is an interrogation tool masqueraded as ‘scientific test’ in which the extraction imperative is secreted under the veil of technological progress. Its purpose and sole potential are not to detect truth, but to enable interviewers to extract confession statements at the price of rationality and legitimacy (see Kotsoglou, 2021; Kotsoglou and Oswald, 2021).

As a result, polygraphic data are uninterpretable (Biedermann and Kotsoglou 2022), i.e., their probative value cannot be assessed as regards truthfulness or the existence of countermeasures. Even (experienced) polygraph operators make clear that polygraphy is primarily ‘a psychological procedure and only secondarily of a physiological nature’ (Abrams, 1978: 178). Relatedly, recent FOI research has revealed that according to the ‘National Polygraph Examination Policy (Standard Operating Procedure)’, ‘the focus of the polygraph process is not solely on whether the subject shows significant responses on a test, but on facilitating admissions on intelligence disclosures that assist in the assessment and supervision of offenders, or the ongoing risk management of a suspect’. ¹²

Since social reality is complex and every conversation unique, the polygraph interview cannot be standardised. Early on in the history of the polygraph it became clear that ‘wide divergence’ in the structure and content of the respective interview is inevitable due to the ‘widely varying types of questions, examiners, and examinees’ (Office of Technology Assessment, 1983: 11). As the UK Ministry of Justice/Probation Service admitted following an FOI request: ‘there are no set questions that come with this [polygraph] test. Each test is created and based on the specific licence conditions of the individual’—unsurprisingly. What is more, several police forces replying to FOI requests noted that each test is bespoke to the individual, which is equally unsurprising in view of the complex nature of historical events and human conversation. ¹³

Expectedly thus, Bryan J observes critically in the context of Prohibition 13 that ‘[t]he remit of any topics to be considered during that testing are not defined, as we consider they should have been’ (at [25]). Yet the interview practice that produces aberrations from the point of view of a methodologically valid scientific test produces consistency from the perspective of an interview calibrated towards extracting disclosures from (apprehensive) interviewees.

The emphasis on utility is a point conceded by the US National Research Council too, when they state that one of the roles of the polygraph test ‘is to influence the conduct of a polygraph interview’. A polygraph interviewer, they note:

who detects what he or she believes to be deceptive responses during the polygraph test normally conducts the remainder of the interview differently than an examiner who sees no signs of deception. Such an examiner may ask more probing questions, do additional charting, shift to a different type of polygraph test protocol, or take a more confrontational attitude in the interview in an effort to elicit an admission or to ‘clear’ the examinee of suspicion. (National Research Council, 2003: 22; emphasis added).

This means that according to its own terms, the polygraph is not an objective scientific procedure but an interrogation tool. The same moment in which the interviewer forms a subjective suspicion, he/she utilises the polygraph to extract a confession statement. That is at best a valuable interrogation technique, provided that all relevant procedural and evidential safeguards are in place (PACE 1984). Either way, the polygraph is not a valid test. For the polygraph interferes with the very object of inquiry.

This brings us to the main issue underpinning this case. Bryan J lamented among other things that ‘[t]he remit of any topics to be considered during that testing are not defined, as we consider they should have been’ (at [25]). Such a specification, however, cannot be done due to lack of standardisation for the polygraph, which is an issue from the viewpoint of foreseeability (Articles 5 and 8 ECHR).

Oppression

We saw above how the judgment accepted that Prohibition 13 was ‘oppressive’ due to the latter's considerable vagueness. I agree with the conclusion, although I submit the use of the polygraph should be declared oppressive for an additional reason too. The remit of oppression, which is partially defined in s. 76(8) of PACE, also includes deception. ¹⁴ The problem, obviously, is that HMPPS are confronted with an unpalatable dilemma:

either to inform the interviewee that the polygraph has no valid scientific basis and is only designed to extract disclosures from apprehensive individuals, significantly thus undermining its utility;

or to make misleading indeed deceptive claims.

Worryingly, the penal system or at least HMPPS opts for the latter. For example, the National Offender Management Service (NOMS) in its leaflet called ‘Mandatory Polygraph Testing—Information for Offenders’ explicitly claims that ‘[y]ou will fail a polygraph examination if the results suggest that you have not told the truth’. This statement is unjustified, indeed misleading. Similarly, the NOMS’ claim in the same leaflet that the polygraph ‘is a device which measures the reactions that the human body gives when we tell a lie’ is also deceptive. The polygraph cannot reliably indicate deception (or indeed any other psychological concept). Both are intentionally false claims constituting oppressive measures.

I submit thus that the judgment dealt merely with the tip of the iceberg. More importantly, it dealt with the vagueness of the Prohibition. The problem, as I showed above, is that the approved rules of the polygraphy profession severely disrupt and violate the procedural and evidential architecture of the penal system. More specifically:

due to lack of standardisation, there is a wide variety of topics that the respective interview will cover. This leads to lack of foreseeability which is required by Articles 5 and 8 ECHR;

due to lack of scientific validity, offender managers have to lie to the offender as regards the characteristics and possible outcomes of the polygraph. This should qualify as oppressive too.

The polygraph pushes thus a sharp needle into the fabric of the penal system, especially vis-à-vis the latter's Rationalist foundations. The more we deploy pseudo-scientific devices, the less foreseeable the administration of the law becomes. Pseudo-scientific devices can in principle not be deemed necessary in a modern penal system. A pseudo-scientific device cannot pass the necessity test as laid out in s. 346 SA 2020. We expect that the Court of Appeal will soon have to grapple with further aspects of the use of the polygraph by HMPPS and UK police forces (Kotsoglou and Oswald 2025).