Arguments against veganism

Common arguments against veganism &
evidence-based ethical and practical responses

The central argument

Veganism is often debated as though it were primarily a question of nutrition, environmental sustainability or personal preference. These are important, but the central ethical question is more fundamental: when is it justified to cause harm to sentient animals for human benefit when reasonable alternatives are available?

Three propositions frame the debate.

First, there is substantial scientific evidence that many animals are sentient and capable of experiencing pain, distress and other negative states. Their interests therefore deserve serious moral consideration.

Second, animal agriculture can impose harms on animals and significant environmental costs. The scale of those harms varies across species, production systems and regions, but they should be evaluated using evidence rather than assumptions about tradition, nature or farming practices.

Third, many people have access to plant-based foods and can meet their nutritional needs on a well-planned vegan diet, provided that essential nutrients—especially vitamin B12—are reliably supplied. This is not equally feasible for everyone, and particular care is required during pregnancy, infancy, childhood and other periods of increased nutritional vulnerability.

These propositions do not, by themselves, prove that every person in every circumstance must be vegan. An ethical conclusion requires an additional principle: avoidable serious harm should not be imposed merely for trivial benefits when reasonable alternatives can meet the same legitimate needs. The strength of the conclusion therefore depends on the seriousness of the harm, the importance of the human interest, the availability of alternatives and the costs of changing course.

Where animal products are not needed for nutrition, survival or another compelling purpose, their consumption requires a justification proportionate to the harm involved. Taste, habit and convenience can explain a practice, but explanation alone is not moral justification.

1. Ethical and philosophical arguments

Whose interests count? Anthropocentrism and speciesism

The ethical case for veganism begins with a question: do animals’ interests matter morally, and if so, how much? The “food chain” and “humane slaughter” arguments often assume that animal interests can be outweighed by ordinary human preferences. A vegan argument challenges that assumption, but should make its own underlying principle explicit.

• Anthropocentrism gives humans a central or overriding place in moral consideration. Some versions hold that only humans possess full moral standing; others recognise animal interests but give human interests priority in conflicts. These are distinct positions and should not be treated as identical.
• Speciesism is commonly used to describe discounting an individual’s interests because it belongs to a particular species. Whether the term is defined neutrally or carries a moral judgement is itself debated (Horta 2010; Jaquet 2019). Calling a practice speciesist does not, on its own, prove that it is wrong; the substantive argument must show why species membership is insufficient justification for disregarding comparable interests.

The core argument. Intelligence, language and moral agency may differ across species, but they are not necessary for having an interest in avoiding pain. Some humans also lack particular cognitive abilities, yet we do not generally conclude that their suffering can be ignored. This challenges attempts to make sophisticated reasoning or language a prerequisite for moral consideration (Singer 1975; Regan 1983).

This argument does not require claiming that humans and other animals have identical capacities, rights or needs. Equal consideration means giving comparable interests appropriate weight, not treating all beings identically. A human may have interests in political participation that a pig does not; a pig can nevertheless have a strong interest in avoiding pain and distress.

Serious objections and qualifications

• Relational views. Diamond (1978) argues that our moral relations with animals cannot be reduced to calculations of interests or capacities. Seeing animals as fellow creatures may ground responsibilities in relationships, practices and forms of life.
• Political approaches. Donaldson and Kymlicka (2011) argue that responsibilities vary with animals’ relationships to human communities, including domesticated, wild and liminal animals. This approach raises questions about care, habitat, membership and coexistence, not only individual suffering.
• Ecological perspectives. Ecosystem-centred approaches may prioritise ecological integrity or species conservation, sometimes in tension with the interests of individual animals—for example, in wildlife management.
• Human priority. One can reasonably argue that human needs sometimes take priority. The vegan response need not deny this; it asks whether ordinary preferences such as taste or convenience justify serious harm when adequate alternatives are available.

Connection to human rights. Human-rights frameworks are designed to protect human beings and need not automatically extend identical rights to other species. The relevant question is narrower: can we defend human protections while also recognising that animals’ capacity to suffer gives us reasons not to harm them unnecessarily? The case for animal consideration does not require equating animal rights with human rights.

Conclusion. Speciesism is a useful challenge to unexamined assumptions, but it is not a substitute for argument. The strongest case is that sentience gives animals interests that deserve moral weight, and species membership alone does not explain why those interests should be disregarded for comparatively minor human benefits.

The food chain and the appeal to nature

The argument: Humans are omnivores or sit near the top of the food chain; eating animals is natural, and nature shows that stronger animals may consume weaker ones.

The response

• Natural does not mean morally justified. Nature contains cooperation as well as predation, disease and violence. A behaviour’s biological occurrence does not establish that it is ethically acceptable.
• Ability is not obligation. The fact that humans can digest animal products does not establish that we ought to consume them in every circumstance. Biological capacity answers a descriptive question, not the moral question.
• Necessity and choice differ. Wild carnivores hunt to survive and do not deliberate about food ethics. Humans can reflect on consequences and, in many settings, choose among nutritionally adequate alternatives. This claim should be qualified: access and needs vary, and not everyone has the same choices.
• The relevant test is justification. If an animal product is necessary to meet a serious need and no reasonable alternative exists, the ethical analysis differs from a case where the main benefits are taste, habit or convenience.

Conclusion. Nature explains why predation exists; it does not settle what humans should do when alternatives are available.

Humane farming and slaughter

The argument: Animals can be raised well, treated humanely and killed painlessly, so eating meat can be ethical.

This objection raises two separate questions: how much suffering an animal experiences during life and death, and whether killing the animal is wrong even if suffering is minimised. These questions should not be conflated.

• Welfare matters. Better housing, handling, transport and effective stunning can reduce suffering. Such improvements are worthwhile even for people who believe killing animals for food is wrong.
• Painless killing does not settle the ethics of killing. If an animal has a life worth living, death deprives it of future experiences and other possible goods. Whether that deprivation is itself a wrong depends on one’s account of animal interests and the moral significance of continued life (Regan 1983).
• The replaceability objection. Some consequentialist accounts argue that if one animal is killed and another is brought into existence with an equally good life, total welfare may not decline. Critics respond that this treats individuals as replaceable containers of experience and may fail to respect an individual’s interest in continuing to live. This is a genuine philosophical dispute, not a question settled by welfare science alone.
• The necessity test. Where adequate alternatives are accessible, the justification for killing an animal for food is weaker if the main benefits are preference and convenience rather than health or survival.

Conclusion. “Humane” production can reduce some harms but does not, by itself, answer whether killing is justified. The strongest vegan argument acknowledges improvements in welfare while asking whether the remaining harm is necessary and proportionate.

Indigenous traditions, culture and practicability

The argument: Veganism disregards Indigenous traditions, cultural identity, subsistence needs and communities that depend on hunting, fishing or pastoralism.

This objection identifies real considerations. Food is not merely a consumer choice; it can be bound up with survival, land, identity, community, sovereignty and ecological knowledge.

• Practicability matters. Veganism is commonly framed as avoiding animal exploitation as far as is possible and practicable. Where people lack reliable alternatives or depend on animal foods for survival, the circumstances differ substantially from those of consumers with abundant options.
• Avoid generalisations. Indigenous peoples and local communities are not homogeneous. Their food systems, values and available choices differ, and outsiders should not assume that one ethical prescription fits all.
• Distinguish subsistence from discretionary consumption. The ethical case for reducing animal products is generally stronger where alternatives are accessible and consumption is not necessary for survival or adequate nutrition.
• Cultural importance is relevant, but not an automatic answer to every question. Traditions deserve serious consideration; like other moral considerations, they may still need to be weighed alongside animal interests, human welfare and ecological effects.

Conclusion. A defensible vegan ethic should account for necessity, access, cultural autonomy and local ecology. It should not use exceptional subsistence cases to avoid evaluating avoidable harms in settings where meaningful alternatives exist.

2. Environmental and biological arguments

“Plants feel pain”

The argument: Plants respond to stimuli, so harvesting them causes suffering and plant-based diets are no better than meat-based diets.

The response

• Response is not the same as experience. Plants detect and respond to light, damage, chemicals and other environmental signals. These abilities demonstrate biological responsiveness, but do not establish subjective pain or conscious experience.
• Current evidence. Plants lack neurons and the nervous-system structures known to support pain experience in animals. Taiz et al. (2019) review the evidence and conclude that claims of plant consciousness are not supported by current scientific findings. This is a strong evidence-based conclusion, not a logical proof that no future evidence could change our understanding.
• Keep the terms precise. “Cognition,” “signalling” and “information processing” can describe complex biological functions without implying conscious experience. Claims about plant sentience should be judged by evidence for subjective experience, not by the mere presence of responses to stimuli.
• The resource comparison. Even if future evidence gave plants greater moral significance, it would not follow that eating animals causes less harm. Animals commonly require feed, and producing animal foods can involve growing crops that humans could otherwise eat. However, many livestock consume grass, crop residues or other materials not suitable for human consumption, and some grazing land cannot readily grow crops (Mottet et al. 2017). The comparison must reflect the particular production system rather than a universal feed-conversion slogan.

Shepon et al. (2018) modelled the opportunity cost of animal-based diets in the United States and estimated substantial potential food-energy gains from replacing animal-based foods with plant alternatives. This is a model-based estimate under stated assumptions, not direct proof that every plant-based system is superior in every respect.

Conclusion. Plants’ responsiveness is not evidence that they feel pain in the morally relevant sense. And even if plant interests were given greater weight, the relevant comparison would be the total harms and resources involved in realistic food systems, not simply whether plants are harvested.

Crop-harvest deaths and field deaths

The argument: Crop production kills mice, birds, insects and other animals, so veganism is hypocritical or causes more animal deaths than meat eating.

This is one of the more serious objections because crop production does cause collateral wildlife harm, and the numbers are difficult to estimate reliably.

• A genuine empirical uncertainty. Davis (2003) argues that some diets including grazing animals could cause fewer animal deaths than diets relying heavily on crops. Matheny (2003) challenges this by considering the land and crop requirements of animal agriculture. Fischer and Lamey (2018) review the limitations of field-death estimates and the ethical questions they raise.
• Death counts are not the same as harm measures. Estimates can differ depending on which animals are counted, whether deaths are directly observed or inferred, whether insect deaths are included, and whether the measure is deaths, suffering or population-level effects.
• Animal products also involve indirect impacts. Feed production, land conversion, fencing, pest control and grazing management can affect wildlife. The relevant comparison must include these impacts rather than counting crop deaths only for plant foods.
• Intention and foreseeability matter, but do not settle everything. Deliberately breeding animals into systems that predictably kill them is morally different from unintended deaths during food production. However, foreseeable indirect harm remains morally relevant and should be reduced where feasible.
• Practical response. Support farming practices that reduce wildlife deaths and suffering, including integrated pest management and more careful crop-protection methods, while recognising that these approaches also have costs and trade-offs.

Conclusion. Veganism does not guarantee zero animal deaths. The field-deaths objection demonstrates the need for harm reduction across all food systems; it does not, on the available evidence, establish that animal agriculture is generally the least harmful option.

Regenerative agriculture and grass-fed beef

The argument: Well-managed grazing mimics natural ecosystems, improves soil, stores carbon and can be better for the environment than plant monocultures.

Some grazing systems can improve soil condition, support particular habitats or use land that is unsuitable for cropping. Those possibilities should not be dismissed. But evidence from one farm or ecosystem cannot automatically be generalised to global production.

• Assess the whole system. A credible comparison should include methane, nitrous oxide, feed and fertiliser inputs, land requirements, soil-carbon change, water impacts, biodiversity and the previous and alternative uses of the land.
• Soil carbon is finite and can be reversed. Improved management may increase soil carbon in some settings, but storage has limits and can be lost through drought, tillage, fire or later management changes (Powlson et al. 2011). Methane emissions continue while ruminants are being raised, so temporary carbon gains do not automatically neutralise ongoing emissions.
• Land has an opportunity cost. Livestock use a large share of agricultural land while contributing a smaller share of global food energy. Poore and Nemecek (2018) document the high land requirements and wide variation in the environmental impacts of animal products. Hayek et al. (2021) estimate that restoring some land currently used for livestock to native vegetation could offer substantial climate benefits, though the result depends on where restoration occurs and what ecosystems can recover.
• Grazing land is not all interchangeable. Some land cannot support crops and may be best managed with grazing or conservation. That is a reason for context-specific land management, not evidence that grass-fed beef could replace current global animal-product consumption without substantial trade-offs.
• Regenerative labels need evidence. “Regenerative” is not a single standardised production method. Claims should be evaluated against measured outcomes, credible baselines and transparent accounting, not branding alone.

Conclusion. The strongest argument is not that grazing never has ecological benefits. It is that those benefits must be demonstrated for the specific system and compared with realistic alternatives, including ecosystem restoration. Evidence of local improvement does not by itself establish a scalable climate solution.

Almonds, avocados, quinoa and other “vegan staples”

The argument: Foods associated with vegan diets use excessive water, cause deforestation or rely on exploitative labour.

Some of these concerns are real. Certain crops are associated with water stress, poor working conditions, pesticide exposure or damaging land-use change. A vegan label does not guarantee ethical or sustainable sourcing.

• Compare equivalent products and outcomes. Almond milk can have a higher water footprint than some other plant milks, but plant milks generally have lower greenhouse-gas emissions and land use than dairy milk in life-cycle comparisons (Poore and Nemecek 2018). Water impacts also depend on whether water is rain-fed or irrigation water in a water-stressed basin; a single global water-footprint number can obscure this difference (Mekonnen and Hoekstra 2012).
• No particular crop is required. Vegan diets do not require almonds, avocados, quinoa, imported berries or specialty meat substitutes. Beans, lentils, peas, oats, grains, potatoes, seasonal produce and other locally available foods can form the basis of many diets.
• Labour concerns deserve direct attention. The correct response to exploitation in crop production is not to dismiss it because the crop is plant-based. It is to support traceability, labour protections, fair purchasing practices and better production standards across food systems.
• Quinoa requires context. Bellemare et al. (2018) found that the quinoa price boom did not reduce household welfare in the Peruvian households studied and raised incomes for producers. This finding is specific to the study’s setting and should not be generalised to all producers or all commodity booms.
• Deforestation is a comparative issue. Permanent land-use change for commodities is a major driver of forest loss (Curtis et al. 2018). Cattle and oilseed products have been linked to agricultural deforestation in tropical regions (Pendrill et al. 2019), while soy’s role varies by region and use. Much soy is used as animal feed, so blaming soy consumption on vegan diets alone is misleading.

Conclusion. The existence of harmful plant-food supply chains is a valid reason to improve sourcing, not a reason to assume animal products are better. Compare the actual products, impacts and alternatives, and address labour and environmental harms wherever they occur.

“Humans need animal agriculture” and whole-system modelling

The argument: Livestock provides manure, uses land humans cannot crop, recycles by-products, supports rural livelihoods and supplies nutrients; eliminating it could create environmental or nutritional problems.

This is a serious systems-level objection. Livestock performs functions beyond producing meat, milk and eggs, and a model that removes animals entirely must account for those functions.

White and Hall (2017) modelled a hypothetical removal of livestock from US agriculture and found that the reduction in greenhouse-gas emissions would be modest relative to total national emissions and that some nutrients would become more difficult to supply. The result is important, but it is specific to the model, country, assumptions and counterfactual examined. It does not show that present levels or patterns of animal-product consumption are optimal.

A useful response distinguishes three questions:

1. Is total elimination feasible everywhere? Not necessarily under current food systems, diets and constraints.
2. Can reducing high-impact animal products produce benefits? Many studies find that substantial dietary shifts can reduce environmental pressure, especially where current consumption is high (Springmann et al. 2018; Willett et al. 2019).
3. Does environmental utility settle the ethical question? No. A product’s role in nutrient cycling or land management is relevant, but it does not automatically justify every use of animals. Alternatives and transition strategies must be assessed.

Conclusion. Whole-system evidence cautions against simplistic claims that removing all livestock would solve every food-system problem. It supports a more precise discussion of which animal products, in which places, for which purposes, and at what scale can be reduced or replaced.

3. Nutritional and health arguments

Protein and “incomplete” proteins

The argument: Plant foods lack essential amino acids, are poorly absorbed, and make adequate protein difficult to obtain without meat.

The response

• Protein requirements can be met. Well-planned vegan diets can provide sufficient protein through legumes, soy foods, seitan, nuts, seeds and grains (Melina et al. 2016; Mariotti and Gardner 2019).
• Protein quality matters. Plant proteins can differ in digestibility and amino-acid profile. Cereals tend to be relatively low in lysine, while legumes tend to provide more lysine. This is a reason for dietary variety, not evidence that plant proteins are inherently inadequate.
• Complementary proteins need not be paired at every meal. A varied diet over the course of the day can provide the essential amino acids required. People with higher needs, limited energy intake or restricted food choices may need more careful planning (Young and Pellett 1994).
• Avoid false equivalence. “Protein can be obtained from plants” does not mean every vegan diet automatically supplies enough. Total energy intake, food variety, age, health and individual requirements matter.

Conclusion. Protein adequacy is achievable on a well-planned vegan diet, but it depends on sufficient food and appropriate choices rather than on the vegan label alone.

Vitamin B12: a real requirement, not a reason for denial

The argument: Veganism is unnatural or nutritionally deficient because unfortified plant foods do not provide a reliable source of vitamin B12.

This objection contains an important fact: B12 deficiency can be serious, and vegans need a reliable source through fortified foods or supplements.

• State the requirement plainly. Unfortified plant foods are not reliable B12 sources, and some products described as plant sources contain inactive analogues. Vegans should not assume that fermented foods, seaweed or unfortified foods provide adequate B12 (Watanabe 2007; Pawlak et al. 2013).
• Supplementation is essential, not optional. A vegan diet needs a dependable B12 plan. The dose and form should follow reliable health guidance, and people with symptoms or risk factors should seek professional assessment.
• B12 deficiency is not exclusive to vegans. Malabsorption and low intake also affect omnivores, particularly older adults, but this does not remove the specific need for vegans to obtain B12 reliably (Allen 2009; Stabler and Allen 2004; Green et al. 2017).
• The biological source is microbial. Vitamin B12 is produced by microorganisms. Modern animal production also often relies on supplemented feed or other controlled inputs to ensure adequate animal B12 status. Taking a supplement directly is therefore not inherently less legitimate than obtaining B12 indirectly through an animal.
• “Natural” is not a nutritional test. Iodised salt, fortified flour and vitamin D supplementation are common public-health measures. Whether a nutrient is obtained from a supplement or food does not by itself establish whether the overall diet is healthy.

Conclusion. B12 is a genuine and non-negotiable planning requirement for vegans. It is a manageable requirement for many people with access to reliable supplements or fortified foods, but it should never be minimised or left implicit.

Other nutrients and overall health

A vegan diet requires attention to iodine, calcium, vitamin D, iron, zinc and omega-3 fatty acids, including long-chain DHA and EPA where appropriate. Algal oils can provide DHA and EPA; the best approach depends on individual needs and dietary intake (Melina et al. 2016).

The evidence should be divided into three different questions:

1. Nutritional adequacy: Can the diet supply essential nutrients? In many circumstances, yes, with planning and reliable B12, but adequacy is not automatic.
2. Health outcomes: Are rates of disease or mortality different? Observational studies and reviews report associations between some plant-rich dietary patterns and lower risks of certain conditions, but confounding and differences in overall diet quality limit causal conclusions.
3. Risks and trade-offs: Are there outcomes that warrant attention? Yes. The EPIC-Oxford cohort reported higher fracture risks among vegans than meat eaters (Tong et al. 2020). In a separate analysis, vegetarians had lower ischaemic heart disease risk but higher stroke risk; vegan-only estimates were not statistically significant (Tong et al. 2019). These findings should be reported accurately rather than selectively.

A meta-analysis found associations between vegetarian diets and lower incidence of ischaemic heart disease and total cancer, and between vegan diets and lower cancer incidence, though the vegan evidence was based on fewer studies (Dinu et al. 2017). Such associations do not establish that veganism alone caused the outcomes.

Conclusion. The strongest defensible claim is that a well-planned vegan diet can be nutritionally adequate and may offer health benefits, but it requires attention to key nutrients and does not guarantee better health than every well-planned omnivorous diet. The paper should avoid both blanket health claims and blanket claims of nutritional inadequacy.

Soy and hormones

The argument: Soy contains phytoestrogens that cause hormonal imbalance or feminising effects in men.

The response

• Distinguish phytoestrogens from human oestrogen. Soy isoflavones interact with oestrogen receptors differently from the body’s own oestradiol; they are not simply equivalent to taking oestrogen.
• Clinical evidence. Meta-analyses of clinical studies have not found clinically meaningful effects of soy foods or isoflavones on testosterone or oestrogen levels in men (Hamilton-Reeves et al. 2010; Messina 2010).
• Keep health claims measured. Soy can be a useful source of protein, but claims that it prevents particular diseases should be calibrated to the strength and type of evidence (Messina 2016).
• Avoid whataboutism. Pointing to hormones in dairy does not answer the soy question. Each claim should be evaluated on its own evidence.

Conclusion. The available clinical evidence does not support the broad claim that ordinary soy consumption causes hormonal feminisation in men.

Pregnancy, infancy and childhood

The argument: Vegan diets are unsafe for pregnant people or children because they can cause nutrient deficiencies or impair growth.

This issue deserves a higher standard of caution than general adult nutrition because nutritional deficiencies during pregnancy, infancy and childhood can have serious consequences.

• Adequacy depends on planning. A vegan diet in these life stages requires reliable B12 and careful attention to energy, protein, iron, iodine, calcium, vitamin D, zinc and omega-3 fats. Needs and risks differ by age and circumstance.
• Professional guidance matters. Families should work with a qualified clinician or dietitian experienced in paediatric or maternal nutrition. Growth, intake and relevant nutrient status should be monitored as advised.
• Evidence and professional guidance differ. The Academy of Nutrition and Dietetics’ 2025 position paper addresses nonpregnant, nonlactating adults and concludes that appropriately planned vegetarian and vegan diets can be nutritionally adequate; it should not be cited as direct evidence for pregnancy or childhood (Raj et al. 2025). The German Nutrition Society’s 2024 position says the evidence is insufficient for a clear recommendation for or against vegan diets in vulnerable groups and recommends specialist counselling. A 2024 systematic review of vegan diets in pregnancy found only six eligible studies, underscoring the limited evidence base (Klug et al. 2024; Meulenbroeks et al. 2024). Earlier professional guidance is also available, but its date and scope should be made explicit (Melina et al. 2016; Baroni et al. 2019).
• Do not overstate certainty. The fact that a diet can be adequate under appropriate conditions does not show that it is easy to implement or equally suitable in every household or medical circumstance.

Conclusion. It is not accurate to declare all vegan diets inherently unsafe for children or pregnancy, nor is it responsible to dismiss the risks. Reliable supplementation, professional guidance and monitoring are central to the argument.

4. Socio-economic and development arguments

Elitism and cost

The argument: Veganism is a privileged lifestyle because specialty products and supplements can be expensive and affordable food options are not equally available.

This objection is partly correct. Vegan diets do not require premium products, but food affordability and access vary sharply across countries and households.

• Staple foods can be inexpensive. Beans, lentils, peas, grains, oats, potatoes and seasonal or frozen vegetables can form the basis of many plant-based meals. Mock meats and organic produce are optional, not defining requirements.
• Affordability is location-dependent. Modelling suggests plant-based dietary patterns can be cheaper than current diets in high-income and upper-middle-income countries, but may be more expensive in low-income and lower-middle-income settings at current prices (Springmann et al. 2021). Hirvonen et al. (2020) found that even a nutritionally desirable reference diet was unaffordable for many people in low-income settings.
• A cheap staple list is not a complete answer. Nutritional adequacy depends on access to a variety of foods, reliable B12, and other nutrients. Time, cooking facilities, transport, disability and local food availability also matter.
• Subsidies need precise evidence. Governments support different parts of food systems in different ways. Claims that animal products are artificially cheap should identify the policy, location and effect rather than generalise from selected subsidies.
• Practicability is ethically relevant. People facing poverty, food insecurity or medical constraints cannot fairly be judged as though they had the same options as affluent consumers.

Conclusion. Vegan eating can be affordable in many settings, but it is not universally cheaper or equally accessible. A credible ethical position asks more of people where meaningful choice exists and supports the social conditions that make lower-harm choices practical.

Livelihoods and food security in lower-income settings

The argument: Livestock supports smallholders and pastoralists, supplies nutrients, uses non-arable land and contributes to food security; reducing animal products could harm vulnerable communities.

This is a substantial objection, not merely a distraction. Livestock supports livelihoods for many households and can supply nutrients where dietary diversity is limited (Herrero et al. 2013; Thornton 2010).

• Context matters. The role of animal foods varies with local ecology, income, food availability, disease burden, livelihoods and nutritional status. A high-income consumer’s discretionary consumption is not equivalent to an animal product that fills a serious nutritional gap in a food-insecure setting.
• Target reductions proportionately. The case for reducing high-impact animal products is generally strongest in populations with high consumption and accessible alternatives. Modelling of dietary changes often finds substantial potential benefits when shifts are concentrated in high-consumption regions (Springmann et al. 2018).
• Transition policy matters. Changes to food systems should protect workers and smallholders, respect pastoralist communities, and consider replacement livelihoods and nutrition. Ethical advocacy should not treat people whose livelihoods depend on livestock as disposable obstacles.

Conclusion. The evidence supports context-sensitive dietary change, not a simplistic global prescription that ignores nutrition, livelihoods or local ecology.

“Most people who try veganism give it up”

The argument: Many people who try veganism do not maintain it, so veganism is unrealistic or unsustainable.

• Adherence is a practical question, not a complete ethical rebuttal. Difficulty maintaining a practice does not establish that its underlying ethical reasoning is wrong. It does, however, matter when designing effective recommendations.
• Interpret survey data carefully. Estimates of how many people stop following a diet depend on definitions, sampling, duration and whether the data are peer-reviewed. Avoid broad claims such as “most people” without a reliable, clearly described source.
• Support improves feasibility. Affordable food, cooking skills, social support, appropriate nutrition advice and accessible options can help people maintain dietary changes.
• Incremental change can still matter. Reducing animal products may deliver benefits even when a person does not adopt a fully vegan diet. This practical point need not abandon the ethical case for veganism as an aspiration or principle.

Conclusion. Adherence data should inform practical support and realistic transition strategies, not be treated as a direct verdict on the ethical argument.

5. The strongest opposing evidence—and what it establishes

The following objections deserve serious engagement because they challenge important parts of the case rather than merely repeating familiar misconceptions.

Whole-system modelling

White and Hall (2017) found that eliminating livestock from US agriculture would reduce greenhouse-gas emissions only modestly at the national level and would create gaps in some nutrients under their model assumptions. This cautions against claiming that removing every farm animal would automatically optimise the food system. It does not establish that current consumption is optimal, or that reductions in high-impact animal products cannot provide benefits.

Least-harm reasoning and crop deaths

Davis (2003) raises a genuine question about whether some animal-inclusive systems might cause fewer field-animal deaths than crop-based systems. The empirical estimates remain uncertain, and the answer depends on what is counted and how systems are compared. The paper should not claim that the issue is conclusively settled. It should argue for the best available whole-system comparison and for reducing avoidable harm in all food production.

Mostly plant-based versus strictly vegan

Environmental and health benefits may be achieved by substantial reductions in animal products without everyone becoming vegan. Environmental models often examine dietary patterns rather than the moral status of animal use (Springmann et al. 2018; Willett et al. 2019). The marginal environmental benefit of moving from a mostly plant-based diet to a fully vegan one can be smaller than the benefit of moving away from a high-animal-product diet. The case for strict veganism therefore rests partly on ethical concerns about exploitation and killing, not environmental outcomes alone.

Nutritional risks

Nutritional adequacy is achievable for many people on a well-planned vegan diet, but deficiency risks are real, particularly when B12 is not supplied or when diets are restrictive, energy intake is inadequate, or needs are elevated. A responsible argument must state the requirements and avoid presenting potential adequacy as automatic adequacy.

Different ethical theories may reach different conclusions

Consequentialists, rights theorists, relational ethicists and ecocentric thinkers do not share a single theory of what matters morally. The paper should explain why it finds the vegan conclusion persuasive rather than imply that all serious ethical frameworks necessarily endorse it. The most robust case combines a clear ethical principle with evidence about the harms and alternatives in specific contexts.

Conclusion

The case for veganism is strongest when it avoids claims that are broader than the evidence. It does not need to show that every animal product is equally harmful, that vegan diets are universally healthier, that crop production causes no animal deaths, or that every community can adopt the same diet.

It needs to show why animal suffering and other harms matter morally; why common objections do not automatically justify those harms; and why avoiding animal exploitation is a reasonable response where practical, nutritionally adequate alternatives exist.

The central principle is simple but not simplistic: imperfection is not a sufficient defence of avoidable harm. Ethical choices should be judged by the best available evidence, the alternatives realistically available, the seriousness of the interests at stake and the consequences for everyone affected. This supports a strong case for reducing and, where practicable, avoiding animal exploitation—while remaining honest about uncertainty, nutritional responsibilities and social constraints.