Haydon v Rex: “Likely” in s.2 Explosive Substances Act 1883 means “could well” (real risk), and mens rea is limited to causing an explosion

1) Introduction

Haydon v Rex ([2026] EWCA Crim 12, 21 January 2026) is an important Court of Appeal (Criminal Division) decision clarifying two core elements of the offence under section 2 of the Explosive Substances Act 1883: (i) what “likely” means in the phrase “likely to endanger life or to cause serious injury to property”, and (ii) the scope of the offence’s mental element (“unlawfully and maliciously”).

The appellant, Matthew Haydon, was convicted after a 10-day jury trial at the Central Criminal Court of causing an explosion by an explosive substance. The factual setting was domestic and experimental: Haydon kept chemicals and conducted experiments in an upstairs room converted into a makeshift laboratory. An explosion seriously injured him, damaged windows, and triggered evacuation and EOD involvement. The prosecution case was that he caused an explosion by an IED/explosive substance of a nature likely to endanger life or cause serious injury to property. The defence characterised him as an accident-prone amateur who did not intend or foresee an explosion, and disputed whether the explosion met the “likely” threshold.

On appeal he advanced three grounds: (1) misdirection on the meaning of “likely”; (2) wrongful admission of prejudicial evidence (photographs and material showing devices/diagrams) under section 78 of the Police and Criminal Evidence Act 1984; and (3) misdirection on mens rea, contending that “maliciously” required foresight not merely of an explosion, but of an explosion of a magnitude that was objectively “likely” to endanger life/seriously damage property.

2) Summary of the Judgment

Appeal dismissed.
“Likely” in s.2 means “could well” (a real risk/real capacity), not “more likely than not”.
The disputed evidence (photographs and materials about devices/diagrams) was admissible; its probative value on knowledge/intent was substantial, and unfair prejudice was manageable, particularly given the Crown did not allege a plan to harm others.
The mens rea for s.2 is confined to intending to cause an explosion or foreseeing the possibility of causing an explosion and continuing to take that risk; there is no requirement to foresee the magnitude of the explosion or the level of harm.

3) Analysis

A. Precedents Cited

(i) Authorities on the meaning of “likely”

The court treated “likely” as a context-sensitive statutory term, drawing heavily on a line of authority recognising that “likely” can range in meaning from “more likely than not” to “may well/could well happen”.

Re Harris Simons Construction Ltd [1989] 1 WLR 368: relied on by the defence for the proposition that “likely” ordinarily connotes probability; Hoffman J’s “Golden Spurs at 5-1” illustration was deployed to argue that “likely” should mean >50% in a criminal statute. The Court of Appeal did not accept that this fixed meaning should be transplanted into s.2 of the 1883 Act; it preferred a purposive, context-based construction.
In Re H (Minors)(Sexual Abuse: Standard of Proof) [1996] AC 563: foundational authority for the “real possibility/real risk” reading. Lord Nicholls’ distinction between “probable” and “a real risk that ought not be ignored” was central. The Court of Appeal adopted the same conceptual move: where the feared harm is grave and protective purpose is paramount, “likely” need not be pegged to a >50% threshold.
Cream Holdings Ltd and others v Banerjee and another [2004] UKHL 44: Lord Nicholls’ statement that “likely” can encompass meanings varying from “more likely than not” to “may well” reinforced the elasticity of the term and legitimised choosing a lower threshold where serious consequences make a rigid probability test irrational.
R v Neil Whitehouse (1999) (Case No.1999/04553) [2000] Crim.L.R.172: particularly influential because it is a criminal case concerning conduct “likely to endanger” (air navigation). Pill LJ’s description of “likely” as “protean” and the conclusion that, where consequences can be disastrous, “likely” means “a real risk, a risk that ought not be ignored” provided a close analogue to explosives: rare but catastrophic outcomes justify a “real risk” threshold.
Chief Constable of Lancashire v Potter [2003] EWHC 2272 (Admin): deployed to show that context can pull the other way. Auld LJ favoured “more probable than not” for ASBO-type restrictions, distinguishing contexts where the “riskiness” and gravity are typically lower than child-protection or catastrophic-risk regimes. The Court of Appeal used Potter to underscore—rather than undermine—the need to choose meaning by statutory purpose: s.2 is aligned with catastrophic-risk protection, not ordinary civil restraint.
SCA Packaging Ltd v Boyle [2009] UKHL 37: reinforced that predictive judgments (especially about outcomes and risk) often cannot sensibly be reduced to >50% and may instead be framed as “could well happen”. The Court of Appeal treated s.2 as involving assessment of an explosion’s capacity and risk, not proof of a more-likely-than-not outcome.
R v Marcus [2013] NICA 60, R v Christine Connor [2021] NICA 2, and R v Paul Campbell [2022] NICA 42: although not binding, these Northern Ireland s.2 authorities were highly persuasive because they interpret the same statutory language. Each adopted a “could well happen” threshold, focusing on the potential of the explosion/device rather than actual consequences. The Court of Appeal (England and Wales) endorsed that approach and agreed with the policy logic articulated in Marcus: Parliament would not rationally criminalise explosions only when the risk edges above 50%.
R v Thacker (Edward) [2021] EWCA Crim 97: considered but distinguished. Thacker involved section 1(2)(b) of the Aviation and Maritime Security Act 1990 and the need for uniform interpretation across contracting states. The court in Haydon treated Thacker’s discussion of “likely” as context-bound and not determinative for s.2, particularly given the different statutory language and interpretive constraints in Thacker.

(ii) Authorities on “malice” and the scope of mens rea

Ground 3 required the court to identify whether “maliciously” in s.2 extends to foresight of the scale of explosion/harm, or is confined to intent/recklessness as to causing an explosion.

G [2003] UKHL 50: cited to articulate modern recklessness as awareness of a risk and unreasonableness in taking it (in the circumstances known). The court used it to anchor the recklessness limb: foresight of the possibility of an explosion and continuing anyway.
Cunningham [1957] 2 QB 396: a classic “malice” case. The appellant argued that Cunningham required foresight of the particular kind of harm. Haydon reads Cunningham (in line with later authority) as not requiring foresight of the degree of harm, and as compatible with treating the endangerment/serious property injury as an objective element once “malice” (intent/recklessness) as to the act is proved.
Mowatt [1968] 1 QB 421 and R v Savage, DPP v Parmenter [1992] 1 AC 699: together confirm that “maliciously” can require foresight of some harm, but not the precise degree (e.g., not necessarily grievous harm). Haydon analogised: requiring foresight of the explosion itself is sufficient; the statutory seriousness is handled by the objective “nature likely to…” component.
R v Thacker (Edward) [2021] EWCA Crim 97 (again): used for contrast. Thacker required intent as to disruption with knowledge/intent that endangerment would be likely “in some shape or form”, but Haydon held that the different wording and statutory scheme prevent importing that higher mens rea into s.2 of the 1883 Act.

B. Legal Reasoning

(i) Ground 1: “Likely” = “could well”, not “more likely than not”

The Court of Appeal’s reasoning is distinctly purposive. Section 2 is directed at protecting the public from grave harms arising from the inherently dangerous and unpredictable use/manufacture/storage of explosive substances. In that setting, a strict probability (>50%) construction would create an arbitrary and under-inclusive boundary, exempting conduct that presents a substantial and unacceptable risk of death or serious property injury merely because it is assessed at, say, 49%.

The court accepted that “likely” is variable in ordinary and legal usage, and that statutory context is decisive. It expressly agreed with the trial judge’s point (echoing Marcus) that Parliament cannot sensibly have intended to criminalise an explosion that is “just” above 50% likely to endanger life, but not one that “might well” do so just below 50%. Accordingly, the jury direction—“could well cause danger to life or serious damage to property”—was upheld.

(ii) Ground 2: Admission of evidence under s.78 PACE 1984

The challenged material comprised photographs and content showing devices/diagrams and items consistent with a developed interest and capability in explosive construction. The defence characterised the appellant as a “foolish and chaotic amateur” and framed the incident as accidental.

The trial judge admitted the material as probative of the appellant’s knowledge, interest, and “know-how” relevant to whether the explosion was accidental and to whether he intended an explosion or foresaw the possibility of one. The Court of Appeal endorsed that evaluation: the evidence went to expertise and awareness, not to proving a plan to harm others. Crucially, the jury were told the case was not advanced on the basis that he planned to use devices against others—reducing the risk of unfair prejudice.

The court also signalled that, even had s.78 been engaged differently, the defence stance risked opening the door to rebuttal for “false impression” under section 101(1)(f) of the Criminal Justice Act 2003.

(iii) Ground 3: Mens rea does not extend to magnitude of explosion or degree of harm

The appellant’s argument attempted to “link” the mens rea to the objective dangerousness element: he contended that the prosecution had to prove intent/recklessness as to an explosion of a magnitude likely to endanger life or cause serious property injury. The court rejected this as unsupported by statutory language and inconsistent with the nature of explosives.

The court held: (a) the mental element is satisfied if the defendant intended to cause an explosion or foresaw the possibility of causing one and continued anyway; and (b) whether the explosion was “of a nature likely to…” is a separate objective inquiry for the jury. No foresight of “any degree of resulting harm” is required.

Addressing the defence hypothetical (unknown additional explosive leading to a much larger explosion), the court treated that as potentially relevant to mitigation, not liability. This reflects a policy choice: those who unlawfully risk causing explosions take their victim and outcome as they come, consistent with the unpredictability and escalation risks intrinsic to explosive materials.

C. Impact

Clarified jury direction for s.2 1883 Act: In England and Wales, this decision authoritatively supports directing juries that “likely” means “could well” (real risk/real capacity), aligning with the Northern Ireland s.2 line (Marcus, Connor, Campbell).
Lowered evidential burden on probability (but not on seriousness): The prosecution need not prove the explosion was more likely than not to endanger life; it must prove the explosion’s nature carried a real, non-ignorable capacity to do so—an important distinction in risk-based offences aimed at catastrophic harm.
Mens rea confined to explosion-causation: Prosecutors do not need to show the defendant foresaw the scale of the blast or the degree of harm. Defendants cannot avoid liability by claiming they expected only a “small” explosion if the explosion caused is objectively of a dangerous nature.
Evidence of technical knowledge and prior experimentation: In homemade explosives cases, courts are likely to admit contextual evidence of equipment, notes, diagrams, and relevant media where it goes to knowledge/intent/recklessness, especially when the defence narrative is “accident/naïveté”.
Sentencing and mitigation: The court left open that unforeseeably extreme consequences may mitigate sentence, even if they do not negate guilt—an important practical lever in cases involving unexpected escalation.

4) Complex Concepts Simplified

“Likely”: Not a fixed mathematical test. Here it means the explosion could well endanger life/seriously damage property—a real risk, not necessarily >50%.
Actus reus vs mens rea: The actus reus is what was done (causing an explosion by an explosive substance, of a dangerous nature). The mens rea is the required state of mind (intending an explosion or foreseeing the possibility of one and taking the risk).
Objective assessment of danger: The jury evaluates the explosion’s “nature” (capacity/potential) objectively—what it could do—not only what it actually did.
“Maliciously”: In this context, it does not mean “spiteful”. It refers to intention or recklessness in the relevant legal sense (here, as to causing an explosion).
Section 78 PACE 1984: Allows exclusion of prosecution evidence if admitting it would adversely affect trial fairness. The court here held fairness was not undermined because the evidence was strongly relevant and its prejudicial risk was controlled by how the case was put to the jury.
Section 101(1)(f) Criminal Justice Act 2003: A mechanism permitting bad character-type material to correct a false impression created by the defence.

5) Conclusion

Haydon v Rex strengthens and clarifies the operation of section 2 of the Explosive Substances Act 1883 in two ways. First, it confirms that “likely” means “could well” in this statutory setting—reflecting Parliament’s protective purpose against low-frequency, high-consequence risks inherent in explosives. Second, it confines the offence’s mens rea to intention or recklessness as to causing an explosion, leaving the dangerousness of the explosion’s “nature” to an objective jury assessment.

The result is a coherent risk-based framework: those who intentionally or recklessly cause explosions cannot evade liability by arguing about sub-50% probabilities or claiming they only foresaw a smaller blast, where the explosion caused was objectively capable of endangering life or seriously damaging property.