The cage and the cause: Why captivity costs more than zoo conservation delivers
Zoos justify keeping wild animals in captivity by pointing to conservation. The argument deserves a hard look, because the costs of captivity are certain and fall on individual animals, while the conservation payoff is real but narrow.
What captivity does to the animals
The best-documented case is the elephant. A 2008 study in Science by Ros Clubb, Georgia Mason and colleagues analyzed records for more than 4,500 elephants. The zoo records represented European populations from 1960 to 2005; comparisons used a protected population of African elephants in Kenya’s Amboseli National Park and Asian elephants working in Burmese timber camps, not wild Asian elephants. The authors reported shorter median lifespans for zoo elephants in those historical data: 16.9 years for female African zoo elephants versus 56 years in Amboseli, and 18.9 years for Asian zoo elephants versus 41.7 years in the timber-camp comparison group. They also reported high infant mortality among Asian zoo elephants and poorer adult survivorship among zoo-born than imported animals. These comparisons raise serious welfare concerns, but they do not directly measure all aspects of welfare or establish the outcomes of current management at every zoo.
A 2010 review of zoo-elephant welfare found stereotypic behaviour—repetitive rocking, swaying or pacing—in a substantial share of the elephants studied. Estimates varied by method and observation period: one sample observed around the clock found the behaviour in 68% of elephants, while daytime-only observations produced lower estimates. Such behaviour is a recognized welfare indicator, though its presence and frequency do not by themselves describe an animal’s full welfare state. Clubb has argued that zoos cannot sustain their elephant populations without importing new animals from the wild.
The same team’s earlier work on wide-ranging carnivores reported similar patterns: the more a species roams in the wild, the more likely it is to show stress behaviours and high infant mortality in captivity. The pattern suggests that some wide-ranging species are especially difficult to accommodate in captivity. It does not show that enclosure design and management are irrelevant, or that every species responds alike. For animals whose natural lives depend on extensive movement and complex social choices, even improved enclosures may leave important needs unmet.
What the conservation case rests on
The defence has real substance. A 2011 analysis estimated that zoo collections represented about 15%—roughly one in seven—of threatened terrestrial vertebrate species. Representation in collections is not the same as a managed breeding program or a contribution to recovery. Captive breeding contributed to the recovery of the Arabian oryx, California condor, black-footed ferret and Przewalski’s horse. A 2011 survey-based paper also reported that the global zoo and aquarium community spent about $350 million a year on wildlife conservation. That is a reported estimate from that period, not a current independently audited total.
These are not trivial achievements, and a fair account has to credit them.
Why the conservation case falls short
The successes are few. An early tally credited conservation breeding with a role in 16 species’ improved status on the IUCN Red List. Critics challenged it, and an updated analysis cut the number to 13. Even the celebrated cases are fragile: the Arabian oryx reintroduction in Oman was hampered by poaching, partly because local communities had not been sufficiently involved.
Zoos mostly keep the wrong animals. Researchers have noted that zoos display mainly large-bodied vertebrates and less-threatened species. Elephants, big cats and primates draw visitors, but most of the species facing extinction are small, obscure and rarely exhibited.
Holding is not saving. Keeping an animal in a collection is not the same as preserving a species, and breeding in captivity does nothing about what is driving extinction in the wild, which is habitat loss, hunting and trade. Reintroduction fails when those threats are not addressed.
The funding does not require the cages. The $350 million figure shows zoos raise money. It does not show that exhibiting elephants and tigers is the way to do it, and field conservation can be funded without keeping thousands of wide-ranging animals in enclosures.
Small commercial zoos and petting zoos
Everything above takes the best case, large, professionally run institutions. Many animal attractions are nothing like that, and the argument is stronger against them.
The Association of Zoos and Aquariums (AZA) listed 238 accredited zoos and aquariums as of 2025. This dated figure should be cited to an AZA source or an archived copy of the AZA list. It estimated about 2,800 exhibits operating under a US Department of Agriculture licence in 2019, fewer than 10% of them accredited. Treat the second figure with care: it counts every licensed exhibitor, including circuses and roadside attractions, so it overstates the number of zoos in the strict sense. The gap still shows how many animal businesses sit outside the accredited sector.
A study that compared USDA inspection records found that non-accredited exhibitors had significantly more non-compliance items than AZA-accredited facilities, including on veterinary care, husbandry and record-keeping. Inspection compliance is only a proxy for welfare, and the effect of accreditation on animals’ actual welfare remains understudied, but the direction is clear.
The conservation record described earlier chiefly concerns coordinated programs and large institutions. A small entertainment business should not claim those sector-wide benefits as its own unless it can show what conservation work it supports. That makes the conservation justification weaker in such cases, though the welfare of its animals still requires direct assessment.
Petting zoos. The best-documented problem is disease, not welfare. Since 2000, petting zoos and farm visits in the US have caused at least 32 publicised outbreaks of E. coli, Salmonella and Cryptosporidium, according to one 2012 review, and a 2012 North Carolina fair outbreak made at least 106 people ill and killed a two-year-old. Cattle, sheep and goats can carry E. coli O157 without showing any sign of illness. CDC guidance notes that animals shed more pathogens when stressed by transport, confinement and crowding, and that most petting zoos exhibit young animals. Studies found handwashing rates of only 25% to 35% among visitors.
Direct research on how petting zoo animals themselves fare is thin, so the welfare case here rests on conditions that CDC guidance links to stress, such as transport, confinement and crowding, and not on measurements of the animals. The ethical objection is also different: these animals are usually domesticated, so the argument about confining wild animals does not apply in the same way. What remains is the use of animals as props for entertainment, with the added cost of risk to children.
The argument in this post is aimed at the strongest version of the zoo case as well as at weaker operators. The evidence about small attractions is less complete, especially on animal welfare, so claims about them should be kept proportionate to what has actually been measured.
The ethics of confinement
Set the numbers aside and a prior question remains: is it right to imprison a sentient animal for purposes that are not its own?
“Imprison” is a defensible word. A zoo animal cannot leave, and it did not agree to stay. Philosophers argue over whether animals have a will in the sense humans do, but the case doesn’t depend on settling that. An animal that paces a fence line, chews bars or repeatedly tests its boundaries is behaving in ways that are hard to read as acceptance. Being able to pursue your own ends is among the interests that matter to creatures that have them.
Three major ethical traditions reach similar conclusions by different routes.
- Rights. Tom Regan argued that animals who are “subjects of a life” have inherent value and may not be treated as mere means to human ends, however good the ends. Regan argued that mature, normal mammals meet this criterion. Applied to zoos, the view implies that a humane enclosure doesn’t fix the wrong, because the wrong is the confinement itself.
- Utilitarianism. Peter Singer’s approach, set out in Animal Liberation, calls for equal weight to be given to comparable interests, whatever the species. Applied to zoos, it doesn’t forbid captivity outright, but it demands that the benefits be large and unobtainable by other means. The earlier sections suggest that for most zoo animals they are neither.
- Justice. Sue Donaldson and Will Kymlicka argue in Zoopolis that wild animals form sovereign communities entitled to protection against invasion and domination of their territory. Applied to captivity, the implication is that capturing them is hard to square with that sovereignty.
Common defences, and replies.
- “They are safe and well fed.” Most people reject that argument for human prisoners. A defender has to explain why species makes the difference, and an answer based on animals lacking any concept of liberty sits awkwardly with the evidence of frustration and stress when they are confined.
- “Captive-born animals never knew freedom.” They still carry the drives to range, forage and hunt, and the evidence doesn’t show them adapting. In the elephant study, captive-born Asian elephants did worse than wild-born ones kept in the same zoos.
- “The species matters more than the individual.” This is the real clash. Conservation values species and ecosystems, while the ethics of captivity concerns individual animals. Philosopher Dale Jamieson argued in “Against Zoos” that there is a moral presumption against keeping wild animals in captivity, so zoo defenders must show important benefits that can be had only that way, and that what zoos teach visitors is a false sense of humanity’s place in the natural order.
Where this leaves the hard cases. Some critics of zoos accept sanctuaries for rescued animals who cannot be released, and short-term emergency breeding for species on the brink of extinction, because in those cases confinement serves the animal or its survival and aims at eventual release. The line these critics draw runs between confinement intended to rescue or rehabilitate animals and confinement for display and entertainment. Whether a particular breeding program is necessary, effective, and feasible without public display is a further empirical question.
The balance
Put the two sides together. The evidence reviewed here gives particular reason for concern about elephants and some wide-ranging carnivores, but it does not establish that every zoo animal has a shortened life or poor welfare. Nor does a species appearing in a collection prove that the collection has helped its recovery. Captive breeding has contributed to some recoveries, while zoos also report contributions through field conservation, research, and visitor engagement. A 2024 meta-analysis found positive average effects of zoo visits and zoo-led programs on conservation knowledge, attitudes, intentions, and reported behaviour; the measured effect on behaviour was small, and the evidence base has limits. Whether these benefits justify captivity for a given species remains an ethical judgment informed by animal welfare and conservation outcomes.
The proposal to shift conservation breeding into a smaller number of specialist centres is therefore a policy position, not a conclusion established by the evidence above. Its feasibility would depend on whether such centres could maintain viable populations, coordinate breeding and research, support reintroduction where appropriate, and secure durable funding without relying on public exhibits. The central ethical question remains whether the conservation and educational benefits of keeping a particular animal on display are substantial, demonstrable, and otherwise unattainable.
Where defenders push back
A fair account should include the strongest counterarguments.
- The elephant data are contested and dated. Zoo scientists dispute the conclusions. Peter Leimgruber of the Smithsonian argues the topic is emotionally driven, and Mason herself concedes that survivorship is not the only measure of quality of life. The core data are from 2008, and zoos have changed practices since.
- Visitor engagement has evidence behind it. A 2024 meta-analysis of 56 studies found positive average effects of zoo visits and zoo-led interventions on conservation knowledge, attitudes, intentions, and reported behaviour. The average effect on behaviour was small, many studies measured near-term or self-reported outcomes, and much of the research came from a few English-speaking countries. This is evidence of potential educational value, not proof that every visit produces lasting conservation action or that display is necessary to achieve it.
- Some species have no other refuge. For species whose habitats are collapsing, captive breeding may be the only practical option left.
- Welfare varies. Species differ widely in how they respond to captivity, so findings for elephants and big carnivores do not automatically apply to all zoo animals.
Sources
- Clubb et al. (2008), “Compromised survivorship in zoo elephants,” Science 322; PubMed record
- University of Guelph news release on the elephant study (2008); Mongabay and Scientific American coverage
- Mason & Veasey (2010), review of zoo-elephant welfare indicators and stereotypic behaviour, Zoo Biology 29; review PDF
- Conde et al. (2011), “An emerging role of zoos to conserve biodiversity,” Science 331
- Gusset & Dick (2011), “The global reach of zoos and aquariums in visitor numbers and conservation expenditures,” Zoo Biology 30; PubMed record; McNally et al. (2024), “A meta-analysis of the effect of visiting zoos and aquariums on visitors’ conservation knowledge, beliefs, and behavior,” Conservation Biology 39; open article; Fa et al. (2014), review of conservation breeding contributions and priorities, Animal Conservation; on ethics, Jamieson, “Against Zoos” in Singer (ed.), In Defense of Animals (1985), pp. 108–117, Regan, The Case for Animal Rights (University of California Press, 1983), Singer, Animal Liberation (1975), and Donaldson & Kymlicka, Zoopolis (2011)
- Accredited versus non-accredited facilities: AZA accreditation figures (AZA institutional statistics; cite a dated source for the figure used); “Does Accreditation by the Association of Zoos and Aquariums Correlate with Animal Welfare Act Compliance?” and Faunalytics summary
- Petting zoos: CDC Compendium of Measures to Prevent Disease Associated with Animals in Public Settings (2004); Werden & Bartlett (2008), Michigan Journal of Public Health; Food Safety News, “The petting zoo problem” (2012)