3.1 Precedents Cited
Ferguson v HM Advocate [2014] HCJAC 19, 2014 SCCR 244
Ferguson was the principal authority and provided the Appeal Court’s framework for three key propositions,
all of which are reaffirmed and sharpened in [2026] HCJAC 5:
-
Mandatory consequence once criteria are met: The Appeal Court repeated that, where the judge concludes the
s 210E risk criteria are met on the balance of probabilities, s 210F means the judge
must make an OLR (the court cites “Ferguson” for this proposition at para [37]).
-
Extended sentence is not a prerequisite “alternative disposal” test: The appellant argued, relying on
Ferguson, that the court should consider whether something “short of an OLR” would address risk. The Appeal Court rejected
any reading of Ferguson that imposes a formal requirement to “consider and discount” an extended sentence before making an
OLR (para [37]). The proper role of extended sentences is more contextual: they may “frequently be in the mind” of the judge,
but do not operate as an intermediate statutory hurdle.
-
Judge’s ultimate responsibility for risk decision: Echoing Ferguson, the court reiterated that the judge is
not bound by the risk assessor’s RMA category; the RAR is influential, but the legal judgment on the statutory criteria is
the court’s (para [38]).
The appellant sought to analogise his case to the “second appellant (Cameron)” scenario discussed in Ferguson,
where an OLR had been quashed. The Appeal Court rejected the comparison (paras [41]–[42]), emphasising that the Ferguson/Cameron
case involved a different personal and offending context (single violent incident with alcohol context, stable family and
employment), whereas Nyiam involved repeated sexual offending against intoxicated/sleeping women, personality disorder traits,
and offending on bail.
3.2 Legal Reasoning
(a) Statutory structure: s 210E (threshold) and s 210F (mandatory disposal)
The court’s reasoning is anchored in the statutory sequence:
-
s 210E defines the risk criteria: whether the nature/circumstances of the offence, alone or
as part of a pattern, demonstrate a likelihood that, if at liberty, the offender will seriously endanger
the public’s lives or physical/psychological well-being.
-
s 210F then requires that if, having regard to the RAR and other information, the court is satisfied
on a balance of probabilities that the risk criteria are met, it shall make an OLR (in
non-compulsion-order cases).
The Appeal Court treated this as leaving no discretion once the risk criteria are found to be met: the OLR is not an “extra”
punitive choice but the legally mandated form of sentence to manage ongoing risk.
(b) Relationship between RMA “Medium risk” and the legal test
A significant contribution of the judgment lies in its careful juxtaposition of:
(i) the RMA risk ratings (high/medium/low) and
(ii) the s 210E statutory criteria.
The court quoted (at para [39]) the then-current RMA definitions, highlighting an important commonality:
both High and Medium begin with the proposition that the “nature, seriousness and pattern”
indicate a propensity to seriously endanger the public. The real difference is that “medium” allows that the
risk may be amenable/manageable and that measures short of lifelong restriction may suffice.
The Appeal Court’s point is doctrinally important: a “medium” RMA rating is not inconsistent with the statutory risk criteria.
The legal question is not whether the offender is “high risk” in RMA terms, but whether the statutory “likelihood” of serious
endangerment is established on the balance of probabilities (paras [38]–[40]).
(c) Evaluating “likelihood”: pattern of behaviour, bail offending, and realism about change
The court upheld the sentencing judge’s scepticism about the assessor’s more hopeful narrative of change. The reasoning turned
on the following evaluative steps:
-
Pattern and repetition: four rape convictions (two in 2020; two in 2025 for 2019 conduct) against broadly
similar fact patterns—complainers intoxicated/asleep, waking to penetration—supported an inference of a persistent modus
operandi and sexual predation (paras [42], [44]).
-
Offending on bail: committing the 2019 rapes while on bail for earlier rape allegations was treated as a
powerful indicator of future non-compliance with restrictive conditions (paras [44]–[45]).
-
Personality traits relevant to enduring risk: antisocial personality disorder diagnosis, near-threshold
psychopathy traits (PCL-R 23), and narcissistic traits were considered highly relevant to the likelihood of serious future
endangerment (paras [42]–[43]).
-
Prison conduct and “protective factors”: the court endorsed the view that “no sexual offending in prison”
carried limited weight because the prison environment removes access to the specific victim profile (intoxicated/sleeping
women) (para [44]). “Protective factors” were also treated cautiously because many would have been present during the
offending period (para [25]).
-
Correct legal threshold: the assessor’s use of “high likelihood” was identified as applying a higher test
than the statute, which requires “likelihood” (para [30], and accepted in submissions at para [35]). The judge was entitled
to adjust for that misstatement when reaching the legal conclusion.
(d) “Looking ahead” to release was not ignored
The appellant argued the judge failed to evaluate risk at the point of future release (a Ferguson theme). The Appeal Court
rejected this, noting that the judge had written submissions on the “look ahead” requirement and expressly considered future
supervision and the unreliability of the appellant’s claimed motivation and compliance (paras [31], [45]).
3.3 Impact
The decision is likely to be influential in Scottish OLR appeals for three reasons:
-
Clarifies the non-determinative role of “Medium risk”: The judgment makes it harder to argue that a “medium”
RMA category, without more, implies that s 210E is not met. Courts may accept “medium” while still concluding that the
statutory “likelihood” threshold is satisfied.
-
Reinforces mandatory nature of OLR under s 210F: Once a judge finds the risk criteria met, the OLR is not a
matter of sentencing preference. This reduces the practical force of submissions framed as “OLR is extraordinary therefore
only if no other disposal will do”, except insofar as they bear on whether the risk criteria are met in the first place.
-
Signals robust judicial scrutiny of optimistic RAR conclusions: The court endorsed a sentencing judge’s
sceptical approach to speculative “amenability to change”, especially where there is demonstrable non-compliance (bail
offending) and entrenched personality traits. Risk assessors’ language and alignment with statutory terms (“likelihood”)
may receive increased attention in future cases.
Practically, prosecutors may view [2026] HCJAC 5 as supporting OLR motions even where the RAR does not label
the risk “high”, provided the factual pattern demonstrates repeated serious sexual endangerment and poor compliance history.
Defence teams, conversely, may need to focus less on the headline “medium” rating and more on undermining the legal inference
of “likelihood” (for example, by evidencing verified treatment engagement, credible insight, and sustained compliant behaviour
in settings that mirror community conditions).