Case Note: Liability Decision in O’Connell v Ministry of Defence [2025] EWHC 2301

Theodore Bunce writes about the decision in O’Connell v Ministry of Defence [2025]. The Claimant, a gunner in the Royal Horse Artillery, suffered a serious shoulder injury during a riding lesson at the King’s Troop when the horse she was riding bucked and unseated her. She pursued a claim in negligence and under section 2(2) of the Animals Act. Deputy High Court Judge Christopher Kennedy KC found for the Defendant on both.

Findings of Fact

The incident occurred during a group riding lesson at in 2015. The Claimant was switched to another horse after struggling to control her first mount. While performing a manoeuvre across the arena, two other recruits turned the wrong way. Believing herself to be in error, the Claimant attempted to turn her horse in the opposite direction, which caused her to lose balance and control. The horse bucked, throwing her to the ground and causing her a significant shoulder injury.

As to the Claimant’s allegations of negligence, the Court rejected claims that her boots were too large and prevented proper use of the stirrups, that the horse was unsuitable for a novice rider, and/or that the instructor failed to act after a prior bucking incident earlier in the lesson. Notable was the Court’s overt reliance on the post-accident record, serving as yet another testimonial to the centrality of contemporaneous documentary evidence when it comes to determining disputed facts.

Reasoning on Animal Act Liability

In short form, section 2(2) of the Animals Act 1971 provides that where damage is caused by an animal that is not of a dangerous species, its keeper is strictly liable if:

(a) the damage is of a kind which the animal was likely to cause, or which, if caused, was likely to be severe;

(b) the likelihood of that damage being caused or being severe was due to characteristics not normally found in animals of the same species, or not normally found except at particular times or in particular circumstances; and

(c) those characteristics were known to the keeper.

The Judge, having remarked (as is by now routine) on the opacity of the legislation, adopted the sequential approach endorsed by Neill LJ in Smith v Ainger (1990) and more recently affirmed in Boyd v Hughes [2025] EWHC 435.

Section 2(2)(a): Was severe injury “likely”?

Having accepted that “likely” in this context meant “reasonably to be expected,” focusing closely on the facts, and adopting the position of an objective spectator to the incident, the Judge concluded that severe injury was not likely. The horse had been trotting, the arena surface was sandy and forgiving, the buck was not particularly violent, and the Claimant was wearing appropriate protective equipment. The Court did not consider how the Claimant landed on the ground, regarding this as being too proximate to the injury actually sustained. Although the Claimant’s injuries were serious, they were the product of misfortune, rather than the reasonably-expected outcome of such an incident. The statistics favoured no severe injury.

Section 2(2)(b): Did the horse exhibit an abnormal characteristic?

The Court had already found when considering the allegations of negligence that the horse had no unusual propensity to buck, but would have found for the Claimant on the second limb of 2(2)(b). Both equestrian experts had agreed that any horse might well buck if it sensed a rider’s loss of balance or control, and/or confusion in the group’s movement. The court accepted that bucking was a characteristic not normally found in horses except in such circumstances.

Section 2(2)(c): Knowledge of the keeper

The Defendant would plainly have been aware that horses can buck when a rider loses balance, or when others in the group move unexpectedly.

Commentary

It is perhaps of note that whilst the Court expressly declined to consider the mechanism of the fall as too proximate to the injury, it also appeared to ignore the novice status of the Claimant in weighing the prospect of severe injury.

Furthermore, even a small degree of reliance on rudimentary statistics regarding injury rates may be too simplistic, and serve to allow the making of a category error; an injury may be statistically unlikely (<50%) but a serious injury still reasonably to be expected given the mechanics involved in unseating and falling. A more sophisticated approach might have brought expert evidence as to biomechanics or epidemiological studies, as opposed to mere observational likelihood.

This decision does stand as a useful demonstration of how the Act is applied in practice, and reinforces just how tightly confined the strict liability under section 2(2) is. Further, it importantly reminds that the statutory test demands a close, fact-specific analysis of the event and the likely severity of injury arising. Generalities about the usual impact of ‘a dog bite’ or ‘a fall from a horse’ will not suffice.