There is a popular belief that electronic containment systems exist simply to keep animals safely within a predefined boundary. That description is not wrong, but it is incomplete. In reality, modern collar-activated electronic pulse systems are used across a wide range of welfare, agricultural, and environmental contexts. They are becoming increasingly popular in use to support rotational grazing across varied pasture types, to manage livestock in terrain inaccessible to man or machinery, and to enable targeted conservation grazing in sensitive ecological environments. In many cases, the boundary itself is established digitally using GPS mapping software and can be instantaneously, remotely adjusted to reflect changing land management requirements.
In public discussion, electronic pulse systems are often framed as though they represent a crude or single-purpose intervention. In practice, they form part of a much wider category of predictable, contingency-based learning systems that allow animals to control their own outcomes through behavioural choice. This principle applies not only to livestock containment systems, but also to handheld electronic training collars sometimes used with dogs, where the purpose is not boundary control but behavioural safety. These systems are used to replace or prevent behaviours that place dogs, livestock, wildlife, cats, motorists, and people at risk. It is important to note that, unlike dogs, there are no handheld electronic training collar systems used for cats. For cats, the only equivalent technology is boundary containment.
At the time of writing (Feb ’26), the Labour government in England has announced plans under their ‘animal welfare strategy’, to conduct a further public consultation regarding electronic training collars for domestic dogs and cats. No date has been specified. Current policy debate has been sharpened by recent public commentary surrounding the use of virtual fencing and collar-activated electronic pulse systems on cattle, particularly within conservation grazing contexts.
In a recent BBC broadcast interview discussing changes to dairy cattle standards carrying the ‘RSPCA Assured’ scheme accreditation, an RSPCA science and policy representative stated confidently, that the charity has banned the use of electronic boundary collars for the herds accredited by their scheme. When questioned in the same interview, the representative went on to explain that:
“a shock is a shock”, but that welfare assessment must consider “the overall picture”, including “the balance of harms versus positives”, particularly where the system prevents animals accessing hazards that would be “much more severely negative” for their welfare. The representative explained that in conservation environments, collar systems utilising electronic pulse “may be justified where it prevents access to more dangerous terrain such as cliff edges or hazardous marshland.”
It is difficult to imagine a more “severely negative” welfare outcome than a slow and painful death.
The broadcast discussion acknowledged that the ethical framework applied was one of balancing potential harms against realised risks, with a focus on overall welfare outcome assessments.
That is an outcome-based welfare model – precisely the welfare-centric model that has historically been rejected by the RSPCA, yet championed by ARDO® when the same discussion is applied to dogs and cats.
The philosophical foundation of modern animal welfare ethics is often traced back to Jeremy Bentham, who wrote in 1789 that the relevant moral question is not whether animals can reason or speak, but whether they can suffer. Both livestock and companion animals are clearly capable of suffering and plenty would argue that they can also ‘speak’ (at least communicate) to those of us prepared to listen. We don’t require laws or lectures to explain what ought to be a basic human empathetic intuition. The difficulty arises when suffering is recognised as morally relevant for some species and contexts, but effectively discounted in others. Equally, acknowledging the ethically proportionate use of the same electronic pulse collar system for one species, based on the system’s capacity for the provision of increased protections (thus elevated welfare), but denying those exact same protections to other animals who might also benefit, based on personal or financial investment, is bordering on Ryder’s aptly-titled ‘speciesism’, where species alone is used to determine an animals’ capacity for suffering – or their right to be protected from it.
When considering dogs and cats, the welfare risks are not theoretical. These are not the “crumbly cliff edges” or “boggy marshlands’ of the mind’s eye. They are real, documented, and recurring.
Across the United Kingdom each year, far too many dogs are needlessly, intentionally destroyed* by veterinary injection following livestock worrying incidents. Destroyed because their behaviour is assessed as posing unacceptable, intractable risk, or killed in road traffic collisions. In addition to elective destruction, legislation allows for dogs involved in livestock worrying to be shot in order to prevent further harm occurring to the animals being worried. While the intention of such fatal action is immediate protection of livestock, death for the dog is rarely instantaneous, yet these proven welfare consequences for the dog are rarely considered within wider policy discussion.
*[The word “destroyed” is far more accurate that “euthanised” or “put to sleep”. These dogs are not mercifully killed to provide a humane release from injury, illness or disease; nor are they waking from their injection. It is important that we use the correct terminology to create the necessary level of understanding. This is about the raw reality of a proven, persistent problem; not a Disney script.]
Equally important is the welfare impact on the innocent victims of such incidents. Livestock worrying events are associated with severe physical injury, fatal trauma, and stress-related losses including miscarriage and abandonment of young. In addition to sheep and cattle, other animals commonly affected by the predatory instincts of dogs include ponies, deer, companion animals, and wildlife species. Then we have motorists and cyclists, whose movement can trigger exactly the same instincts. These outcomes are documented across agricultural reporting, veterinary evidence, species-specific protection groups, road traffic collision reports and policing data relating to livestock worrying.
From a welfare perspective, these harms are neither theoretical nor rare. They arise in situations where control, containment, training or behaviour modification has failed or has been considered unnecessary or unavailable. An ethical framework centred on minimising suffering must consider not only the animal’s reaction to the immediate experience of a controllable and predictable training stimulus such as a brief, harmless electronic pulse, but also the measurable and repeated suffering that occurs when such proactive, preventative interventions are absent.
Research from the VetCompass Programme has demonstrated that mortality in otherwise healthy dogs under three years of age is strongly associated with external causes including road traffic accidents and ‘undesirable’ behavioural outcomes that ultimately result in elective destruction. These were not animals suffering from disease. They were physically healthy animals whose lives ended prematurely due to behavioural or environmental risk exposure. Similar patterns have been observed in international data sets, including Australian research examining early mortality risk factors in healthy companion dogs.
Livestock worrying presents a parallel welfare crisis affecting both livestock and dogs. NFU Mutual has reported that farm animals worth an estimated £2.4 million were severely injured or killed by dogs in 2023, up nearly 30% compared to the previous year. Whist 2024 saw that financial total fall to £1.8 million, the issue remains an issue, and lack of reporting by no means reflects a lack of incidents. Whilst ARDO® are fully aware and sympathetic to the financial aspect of attacks by dogs, it is the associated animal welfare costs that chart our course on the issue.
The National Sheep Association’s 2024 survey revealed that 87% of respondents experienced a dog attack on their sheep in the last 12 months, with attacks becoming more severe and resulting in greater levels of injuries. One respondent reported 44 sheep killed in a single attack.
These events frequently result in severe injury or far from instantaneous death to livestock and often result in healthy, much-loved dogs being legally destroyed following attacks. Again, this is not hypothetical harm. It is realised suffering occurring repeatedly across rural Britain.
The RSPCA’s electronic cattle collar contextual criteria model becomes ethically difficult to defend when we consider the same charity’s blanket rejection of a weaker version of the same electronic pulse, used to train dogs not to worry, attack or kill vulnerable animals, or to keep them safely contained within outdoor areas where conventional fencing or constant supervision or neither possible nor realistic. If a negative stimulus is considered ethically acceptable where it prevents cattle accessing dangerous terrain, or protects grasses and supports ecological land management, then the question naturally follows as to why the same harm-reduction logic is rejected when applied to preventing dog death, cat road mortality, or livestock and wildlife hunting, wounding and killing by uncontrolled dogs?
This straightforward question remains unanswered.
It is also frequently argued that physical fencing is inherently more welfare-appropriate because animals can move away from it, whereas an electronic collar is attached to the animal until removed by the human caretaker. This argument becomes difficult to sustain when we consider the well-documented injury risks associated with barbed wire and commonly used, standard electric fencing**. Such physical fencing can and does cause cuts, entanglement injuries, and occasionally fatal entrapment. Barbed wire for example, is designed entirely around the principles of pain avoidance. Technologically advanced collar-activated systems remove these physical impact trauma risks entirely. Only prolonged wearing of the collars and infrequent checks on the animals by the person responsible, run the risk of pressure-related necrosis.
Consider also, the fact that even where it is possible to erect standard fencing, not only are the costs of erecting and maintaining the fencing prohibitive where pasture extends over large distances, but the fencing is forever prone to damage caused by age related deterioration, weather, hedge growth or animal activity, thus increasing vulnerability and compromising the welfare of the animal the fencing is intended to contain. This issue is not an issue for electronic, geo-fencing systems.
**[A parliamentary question revealed that standard electric fencing which is bought, installed and used without guidance for use with all manner of animals has no legal power limit]
Another distinction cited by the RSPCA during the BBC radio interview is the idea that visual warning cues are inherently more welfare-appropriate than audible cues. Basically, sight is inferred as being a more effective or appropriate means of an animal obtaining valuable, environmental information than sound. In discussing virtual, electronic-collar containment, the speakers moved on to talking about the fact that the cattle containment collars play an audible ‘tune’ before the collar delivers the static pulse. Though repeated experiences with the pairing contingency, the cattle hear the tune as being their cue and opportunity to turn away and successfully avoid the static stimulus.
The RSPCA spokesperson suggested that the absence of a visual boundary may represent a welfare concern, despite the presence of predictive audible warning cues.
No supporting scientific evidence was cited to demonstrate that auditory predictive cues are less effective, less humane, or less welfare compatible than visual cues. Across animal training and livestock management, auditory predictive cues are widely used and well understood by animals. The relevant welfare variables are predictability and controllability via behavioural response, not sensory modality.
Indeed, if we were to follow this assumption to its conclusion, then standard electric fences should be ethically problematic and a matter for the charity to act upon with urgency. They consist of a thin wire, often hard to see and obscured by hedgerows, and animals with impaired or obstructed vision (such as older animals or horses with longer fringes) would be receiving excessive electric shocks*** from the fencing and ought to be removed or their own protection.
Night or low light level grazing would be equally problematic, as would snow / fog / misty morning conditions. In reality, anyone with experience of electric fencing knows that it is the clockwork-like, audible ’ticking’ of the fence as the energiser sends short, high voltage pulses along the wire that serves as the audible pre-warning cue, conditioning the animal to behave appropriately and avoid the boundary shock. Not the sight of the wire. A simple field test, placing livestock that have been contained by electric fencing inside novel fields with a powerless single strand, electric fence permimeter boundary would test the visibility=greater predicatbility theory. The animals should not attempt to breach the wire despite the absence of the audible ticking.
Electric poultry netting is arguably far more visible than electric wire and yet government research reveals that non-target, protected species including amphibians, wild birds and small mammals like hedgehogs have been shown to get tangled, trapped and killed in the netting, unable to free themselves from the automated, continual electric shock.
2021 saw a troublesome time for the RSPB too, after it was revealed in the national press that their own manual relating to the prevention of predation towards birds and their eggs by predators such as badgers, advised deliberately ‘baiting’ perimeter electric netting with honey soaked lures to encourage badgers to hit the fencing and receive shocks. The manual was withdrawn.
***[The word ‘shock’ is deliberately used here instead of the words ‘static pulse’. The high voltage, low-current stimulus received from an electric fence travels through the body of whatever animal (including humans) touches it, down to earth and the fence’s grounding system, completing the circuit. Electronic pulse collars do not travel through the body. Instead, the electrical pulse travels between two points on the collar, which must be touching the skin to create the circuit for the collar to work and the pulse to be felt. This creates a completely different, less powerful sensation to that of an electric fence].
Animals (including ourselves) that can predict and control outcomes via their own behaviours are demonstrably less stressed than animals exposed to random, unpredictable or uncontrollable events.
Research into electronic containment systems in cats has demonstrated that the animals rapidly learn audible, collar-delivered boundary warnings and show no long-term welfare compromise. In some studies, cats exposed to predictable electronic boundary learning systems demonstrated increased confidence and improved performance in subsequent problem-solving tasks. This is consistent with broader behavioural science: predictability supports learning, not helplessness.
It is also important to address the distinction sometimes made between commercial livestock and conservation livestock. In their recent public commentary, the use of electronic containment systems has been opposed by the RSPCA in commercial dairy contexts whilst being accepted in conservation grazing contexts. This distinction raises uncomfortable ethical questions. If suffering is the relevant metric, then it cannot logically matter whether an animal is destined for food production or ecological land management. The animal’s capacity to suffer does not change based on its economic category.
The situation becomes more complex when considering the economic realities surrounding welfare certification schemes. Organisations responsible for welfare labelling must maintain public trust and brand confidence. That is understandable; it is a financial necessity. However, when welfare positions align closely with public emotional perception rather than consistent harm-reduction logic, it risks creating the appearance that image management is being prioritised alongside, or in place of, pure welfare science. Companion animals, due to their emotional proximity to humans, naturally attract stronger public empathy and therefore greater public engagement and charitable support. That reality cannot be ignored when examining policy positions that differ sharply between companion animals and farm animals, or farm animals kept to produce human food bearing the acronym of the decision-maker, and farm animals that do not represent such close, valuable links.
From a purely welfare perspective, the ethical framework should remain constant. If a controlled negative stimulus can prevent greater suffering, then it must be evaluated on outcome rather than emotional perception or species familiarity.
This is not an argument that dogs deserve more protection than cattle. Nor is it an argument that cattle deserve less protection than dogs. It is an argument that suffering should be measured consistently and that consistency should be applied, irrespective of species of investment.
Let me finish with a further ethical conundrum, based on the ‘sight over sound’ and the ’contextual risk calculation’ animal protection model. When a person uses a static pulse collar to train a dog not to worry, attack or kill sheep, it is the sight of the sheep that the dog uses as its predictive stimulus. The dog then performs the appropriate behaviour in any and every situation that the predictive stimulus is present – avoid the sheep – which provides the dog with the necessary control. So here we have the complete contextual protocol prerequisites. A life threatening context; a visual pre-warning predictive stimulus; an appropriate, learned avoidance behaviour; full contextual control; animal welfare promoted or protected.
Every single box ticked. Yet the RSPCA still campaign for a total ban on the tool that makes the whole thing possible.
Perhaps dogs should learn to graze grasslands ……
All factual information contained in this work can be fully referenced.
This piece is written and researched solely by Jamie Penrith.
Please respect the authenticity and the efforts of the author and do not replicate, restructure, rewrite or feed into AI programs under any other name without first obtaining the author’s explicit, written permission.