The study, published in the journal Discover Animals by David C. Cook of the University of Western Australia and Murdoch University, Christina Siettou of University College London, and Rob W. Fraser of the University of Kent, takes an unusually broad view of what is at stake when dog populations are mismanaged. Rather than counting only the costs of kennels and collection services, the researchers define social welfare as the combined well-being of all individuals in society, including both the intrinsic and instrumental value of animals themselves. When free-roaming dogs go unmanaged, the losses ripple outward: wild dog attacks generate medical and veterinary bills, overcrowded shelters impose mental health costs on staff, and rising euthanasia rates create moral distress and opportunity costs for veterinarians and nursing workers. The framework explicitly extends the traditional economic view, in which social welfare depends solely on human well-being, to incorporate the welfare of the animals as well.

To capture the dynamics of the problem, the team constructed a dynamic systems model describing three interacting subpopulations: owned dogs, shelter dogs, and dogs held by registered commercial breeders. Animals move between these groups through purchases, adoptions, and relinquishments, and each subpopulation changes over time through births, deaths, and transfers. The model was simulated over ten years using ordinary differential equations solved numerically in the R programming environment with the deSolve package. Crucially, the flows between populations respond to price. The demand for puppies from breeders and the demand for shelter adoptions are both governed by a price elasticity of demand, assumed to be negative at -0.5, meaning that taxes on breeder purchases or subsidies for shelter adoptions can shift where prospective owners acquire their dogs.

Social welfare enters the model through two channels. Direct losses accrue at a constant rate for every shelter dog euthanised, while indirect losses capture the stress animals endure in crowded facilities, specified as rising at an increasing rate as the shelter population approaches its carrying capacity. Both loss streams are discounted over time using a continuous discounting factor. The researchers set hypothetical values of £50 per euthanised dog and £5 per shelter dog for these parameters, and a sensitivity analysis later confirmed that these two assumptions were the most influential in the entire model, shifting estimated welfare losses by roughly 25 percent when varied by 50 percent.

Against a base case reflecting the status quo, the team simulated five hypothetical policies. Two of them produced sobering, even counterintuitive results. Increasing shelter capacity, simulated as a one-off expansion of the shelter carrying capacity, actually made social welfare slightly worse than doing nothing, producing a cumulative loss £0.6 million greater over ten years. The reason lies in the biology of stress: events that are novel, unpredictable, or outside a dog’s control, such as entering a shelter, trigger the fight-or-flight response and the release of cortisol. Short-term cortisol surges boost energy, but prolonged elevation impairs the immune system and produces anxiety and abnormal behaviours. More kennel space means more dogs housed in stressful conditions, and the model links that directly to greater welfare loss. Increasing the euthanasia rate by roughly 50 percent fared far worse, adding £8.6 million to cumulative social welfare losses compared with the status quo, the poorest result of any scenario.

The economic instruments performed better. A £200 fixed-rate tax on dogs purchased from breeders reduced demand for breeder dogs but left shelter numbers largely unaffected, producing only a negligible change in welfare. A £100 subsidy for shelter adopters, modelled on the German practice of waiving the annual dog tax for shelter adoptions, cut the shelter population by 11,360 dogs after ten years relative to the status quo and reduced cumulative welfare losses by £3.9 million. In the baseline scenario, representing current policies, cumulative social welfare losses over the decade totalled roughly £40.4 million, with annual losses reaching about £5.3 million by year ten.

The clear winner, however, was the DNA database scenario inspired by Greece’s newly legislated framework. Greece, which manages one of the highest per capita free-roaming dog populations in the world, combines compulsory microchipping and registration, a prohibition on euthanasia, strict breeding laws, and mandatory sterilisation with an opt-out route. Owners who decline to sterilise their dogs must submit DNA samples to a national database, meaning any unaltered animal is recorded and its offspring traceable, holding owners accountable for contributing to the future stray population. Non-compliance carries fines ranging from €1,000 for owners who fail to neuter or submit samples to €3,000 for breeders who fail to report or register breeding animals, with enforcement spread across municipal and national police, customs authorities, the coast guard, and forestry services.

In the UK simulation, the DNA policy was represented by raising the effective price of breeder dogs, since breeders pass on the £130 cost of DNA sampling to consumers, lifting the average price from £930 to £1,060; by lowering the natural growth of the owned dog population, as owners sterilise pets to avoid future penalties; and by reducing the relinquishment rate, since sterilisation is associated with fewer surrenders. After ten years, this scenario left 34,000 fewer dogs with breeders, 41,500 fewer dogs in shelters, and 26,210 more owned dogs than the status quo, improving social welfare by £7.3 million cumulatively, or about £1.4 million per year by the end of the simulation. Even when the researchers halved the policy’s simulated effectiveness to reflect widespread non-compliance, it remained the best-performing option, though the cumulative benefit fell to £3.4 million.

The sensitivity analysis revealed which assumptions mattered most. Beyond the two welfare loss parameters, results were sensitive to shelter carrying capacity, the initial shelter population of 250,000, and the euthanasia rate of 10.4 percent drawn from published UK data. Reducing shelter capacity by half produced a 33 percent change in welfare loss, while halving the euthanasia rate lowered losses by 20 percent. Importantly, the researchers note that provided the direct welfare loss per dog euthanised exceeds the indirect welfare loss per shelter dog, which they consider reasonable, the ranking of policies does not change, and the DNA database remains the welfare-maximising choice. The authors acknowledge that the absolute magnitudes are indicative rather than definitive, since the welfare loss values are plausible estimates rather than figures elicited through quantitative research, and they call for future studies to enable a full benefit-cost analysis.

The findings carry an urgent contemporary resonance. In 2024, the government of Türkiye adopted legislative amendments likely to lead to large-scale euthanasia of shelter dogs due to overpopulation, provoking public outcry. The model offers a quantitative warning about such approaches: raising the euthanasia rate from its base of 10.4 percent to 100 percent drove simulated social welfare to minus £135.6 million, a societal loss more than double the status quo. The study also documents the well-established toll that euthanasia takes on the people who perform it, from compassion fatigue to perpetration-induced traumatic stress among shelter and laboratory staff. Against that backdrop, the authors’ central recommendation is straightforward: the UK government should consider a dog-overpopulation policy modelled on Greece’s, because its superiority proved robust both to alternative parameter values and to imperfect compliance. Compulsory microchipping, introduced across the UK by 2016, has already been undermined by owners supplying incorrect or outdated contact details, a cautionary tale about enforcement that any DNA-based successor would need to confront head-on.

Subject of Research: Economic modelling of dog shelter policy reform and social welfare in the United Kingdom

Article Title: Social welfare implications of policy reform for dog shelters in the United Kingdom

Article References: Cook, D. C., Siettou, C., & Fraser, R. W. (2026). Social welfare implications of policy reform for dog shelters in the United Kingdom. Discover Animals, 3(1), Article 99. https://doi.org/10.1007/s44338-026-00258-z

Image Credits: AI Generated

DOI: 10.1007/s44338-026-00258-z

Keywords: dog shelters, animal welfare, social welfare, DNA database, dynamic systems model, euthanasia, sterilisation, policy reform, United Kingdom, Greece, shelter overcrowding, pet overpopulation