We need more 90% vegans
The case for treating veganism, and morality itself, as a spectrum.
TL;DR: There is no moral cliff between vegan and non-vegan because there is no moral cliff between right and wrong. Actions are only better or worse according to how they affect conscious wellbeing, and are therefore best understood as points on a continuous spectrum. Veganism can be a useful rule, identity and commitment device, but it has no value apart from the suffering it prevents.
I’m a bad vegan. I eat honey, whey, gelatin. If a restaurant has nothing on the menu, I often order non-vegan instead of walking 10 minutes elsewhere. And sometimes, I just really, really want some cheese, so I eat it. By most definitions, I am not a vegan at all: the Vegan Society claims that veganism is "a philosophy and way of living which seeks to exclude — as far as is possible and practicable — all forms of exploitation of, and cruelty to, animals for food, clothing or any other purpose.”
I will make the case that “vegan” is a reductive, though sometimes socially useful binary, imposed on a continuous underlying variable: the amount of suffering a person’s choices are expected to produce.
Consider three people:
A eats animal products at every meal.
B eats them once a week.
C never eats them.
Ordinary usage groups A and B together as “non-vegans”, and only C fully qualifies. Yet on the dimension that matters1, B is plainly much closer to C than to A. The final reduction — from once a week to never — still has moral value, but only in proportion to the additional suffering it prevents. It does not acquire any importance because it crosses a semantic threshold.
Let’s imagine applying similar logic to charitable giving. Peter Singer argues that affluent people should continue donating until giving more would require a sacrifice comparable to the harm prevented. We might therefore define a “donor” in terms similar to the Vegan Society’s definition: “A philosophy and way of living which seeks to donate — as far as is possible and practicable — all income beyond what is needed to meet one’s own basic needs.”
Again, consider a set of people:
A donates nothing.
B has taken the Giving What We Can pledge, and donates 10% of their income.
C donates 40%, far exceeding the pledge. They forgo travel, dining out, and expensive hobbies.
D donates 80%: everything beyond the bare minimum to survive. They move to a cheaper city, abandon all luxuries, and live on rice and beans.
On this definition, only D qualifies as a “donor”. It’s self-evident that the relevant variable here is continuous: as the amount donated increases, so does the expected good. Unlike veganism however, with charitable giving we recognise this and rightly praise even partial steps in the right direction2.
Scalar utilitarianism
The underlying problem here is that we often tend to model morality as binary. Actions are right or wrong; permissible or impermissible; innocent or guilty. I make the claim that morality is instead continuous: actions range from the best available option to the worst, according to their expected effects on suffering and wellbeing.
This view is known as scalar utilitarianism, and is associated with Alastair Norcross (although I was pleased to discover that I had arrived at something close to it independently!)
Scalar utilitarianism ranks actions by their expected utility without adding a threshold that separates “right” actions from “wrong” ones. Moral claims are therefore fundamentally comparative: one action is “better” than another because it produces a better outcome, and to say that an action is “wrong” is simply to say that it is worse than an available alternative. In this way, “wrong” and “less good” are synonyms, describing equivalent positions on the same scale.
This also collapses the distinction between causing suffering, allowing avoidable suffering, and failing to create wellbeing. Ordinary morality may classify the first as wrong and the last as permissible; instead we ask how far each choice moves the world up or down the scale. They are just different routes to the same point on the number line.
This does not mean that the routes never matter: someone who deliberately causes suffering may be more likely to go and cause further suffering in the future. Punishing them may deter others, and distinguishing acts from omissions may support useful norms and prevent people from becoming overwhelmed by unlimited obligations. These are real consequences, and belong in the calculation. But once they have been accounted for, there is no additional moral penalty attached to having caused rather than allowed an outcome.
An implication of scalar utilitarianism is that everyone is a moral underperformer, even Alastair Norcross, and Peter Singer. Every day, we choose comfort, leisure, or consumption over preventing suffering. Our psychological limitations may affect which norms are realistic and useful, but they do not create a zone of moral neutrality around our choices. We will all continue to underperform; scalar utilitarianism simply asks us not to conceal this by placing most of our behaviour inside the absolving category of “permissible.”
This way of thinking risks inviting permanent guilt. If every foregone opportunity to help counts as a moral shortfall, perhaps we should work until we burn out; until we become so overwhelmed that we do nothing. Or, worst of all, it might cause us to abandon utilitarianism altogether for a more psychologically tolerable philosophy.
But the scalar view also dissolves the perfectionism that produces this response: if there is no threshold above which we become morally innocent, there is also no threshold below which our efforts become worthless. Doing 50% of the good you could have done is worse than doing 100%, but it is much better than doing none. Failing to do the best does not erase the value of doing better.
Guilt is itself useful only insofar as it helps us act better. The point is to recognise that our choices have consequences, that those consequences come in degrees, and that every genuine improvement matters.
Animal rights vs animal welfare
I’ve made the claim that actions do not possess value independently of their effects on conscious experience. Animal agriculture produces immense suffering, which is why I am (90%!) vegan.
Many vegans instead appeal to “rights” and “exploitation”. They may, for example, refuse to eat meat destined for the bin because they believe they have no “right” to use an animal’s body — even when eating it would affect neither the animal nor future demand.
Utilitarians however claim that exploitation matters only through its consequences: of two worlds with equal suffering and wellbeing, the more exploitative is no worse. However in practice, exploitation can sustain demand and reinforce harmful institutions. Those effects matter, but once they are held constant, no residual wrongness remains.
Nor is killing inherently wrong. Killing is usually bad because it causes fear and pain, and deprives an animal of a worthwhile future. But killing a being that cannot suffer and loses no valuable future would not be wrong simply because a life ended. Likewise, if a farmed animal lived a genuinely good life that would not otherwise have existed, its eventual death would not erase the positive utility of that life. This however does not describe modern factory farming: it is abundantly clear that factory-farmed animals endure lives that are significantly worse than non-existence.
Nevertheless, rights and anti-exploitation principles can be extremely useful heuristics. Calculating the utilitarian consequences of every decision would be exhausting, uncertain and vulnerable to motivated reasoning. “Do not torture people” is a much safer social rule than “Torture people whenever your own calculation suggests it will marginally increase total wellbeing.” Rights are a useful instrument: they matter because systems that respect them tend to produce better lives, but should not be mistaken for a separate moral currency.
Reducetarianism
If you’re vegan, I encourage you to ask yourself why. Is it to preserve a pure and consistent identity, or to improve the material conditions of animals? I hope it is the latter.
If so, the binary rule of veganism must be judged instrumentally. There is a strong case for adopting the simple rule “do not consume animal products”: it reduces decision fatigue, communicates demand, strengthens social norms and creates a useful political identity. The rule can be binary even though the morality beneath it is continuous.
However, I think that overall, the rigidity of the binary does more harm than good. Treating veganism as all-or-nothing can deter people for whom perfection feels unrealistic. It can turn a single lapse into total failure, direct energy towards policing the boundaries of the label and encourage people to optimise for consistency of identity rather than reduction of harm.
The more flexible alternative is usually called reducetarianism. It is usually presented as a psychologically convenient stepping stone towards the real moral achievement of full veganism. I think that this gets the relationship backwards: reduction is not merely a route to the goal; reduction is the goal. Full veganism is one particularly effective way of achieving it, but is a difference of degree, not kind.
The same all-or-nothing instinct appears elsewhere in politics. Reforms are dismissed as capitulation because they leave an unjust system intact: opposing better working conditions because the factory remains capitalist, or nuclear power because it is not fully renewable. The language of “complicity” reinforces this tendency by treating participation in a flawed system as endorsement of the whole. The relevant questions are simpler: does the change reduce harm, and does it make further progress more or less likely?
When thinking about veganism, we should keep two questions separate. First, how much suffering does a behaviour cause or prevent? Second, when we instantiate this behaviour into a social norm, which labels and rules are most productive? The second question matters, but only because it affects the first. There are advantages to making social rules around binaries, but the rule remains a means, not an end.
Wellbeing and suffering are the only moral currency. Labels such as “vegan,” and concepts such as rights and exploitation matter only insofar as they improve that balance. We should therefore judge our norms not by how pure or internally consistent they appear, but by whether they actually make the world a better place.
Imagine how many more people would be almost entirely vegan if they gave themselves the grace to be imperfect — to eat their grandma’s shepherd’s pie, a slice of birthday cake, a drunken kebab. A movement that welcomed those people might contain fewer perfect vegans, but vastly more vegan meals — and therefore far less suffering. The animals don’t care, so neither should we!
Thanks to Jack Douglass for his incisive comments, as per. To @ratlimit for the tweet that started it all. And especially to Alec Harris, for helping me conceptualise these ideas, work through the argumentation, and challenging me in such a thoughtful way. You’ll make an average utilitarian of me yet!
I will assume henceforth that the reader is broadly consequentialist: that actions should be judged by the outcomes they produce. I expect the rest of the argument will not resonate if you reject this premise, but a full defence of consequentialism is beyond the scope of this article.
I do not mean to strawman vegans: many are well calibrated about the value of partial reduction. It seems however that the dominant framing of veganism remains absolutist, a tendency reflected in the Vegan Society’s definition.
While I partially agree that treating veganism as a binary can have negative effects, I think your proposed solution -- lowering the standards vegans hold themselves to -- creates larger problems than it solves. The crux of your argument seems to be that losing some strict vegans is acceptable if a more welcoming, less perfectionist social environment encourages non-vegans to decrease their intake of animal products by enough to increase the total amount of animal suffering avoided. I understand the math you’re trying to do here. I’m just not convinced the tradeoff actually works!
For one, you literally mention that “someone who deliberately causes suffering may be more likely to go and cause further suffering in the future,” not realizing this squarely applies to 90% vegans. If someone who believes consuming animal products is wrong begins deliberately making exceptions, why would we expect those exceptions to remain neatly contained? Habits deteriorate pretty easily once people stop treating them as rules. Eating dairy once while traveling becomes eating dairy whenever traveling is inconvenient, which becomes eating dairy whenever being vegan is inconvenient. At some point, the occasional exception has just become someone’s normal diet.
I think this would also create a pretty serious cognitive dissonance problem. If people routinely do something they supposedly consider immoral, they either have to change their behavior or weaken the moral conviction making them feel guilty. Your proposal makes changing their behavior unnecessary, so I’d expect a lot of people to do the latter. “Eating animal products causes horrible and unnecessary suffering” slowly becomes “avoiding animal products is generally good when it’s convenient.” That may be a more psychologically comfortable belief, but it’s also a much weaker one! If anything, normalizing exceptions gives people a socially acceptable way to gradually stop caring without ever admitting that they’ve stopped caring.
This also weakens the social pressure that keeps people vegan in the first place. Vegans remind each another, both explicitly and implicitly, that consuming animal products is a serious moral failure. That expectation reinforces good habits and makes individual lapses harder to cognitively dissociate from. If basically everyone in the community starts making exceptions, people will feel guilty because they’re failing to live by their stated convictions, avoid conversations that remind them of that fact, and probably become less willing to do outreach. It’s probably harder to persuade someone that eating animal products is seriously wrong while explaining that you still do it frequently.
More importantly, why assume that maintaining veganism as a clear standard prevents non-vegans from moving toward it? People already go through intermediary stages like pescatarianism, vegetarianism, reducetarianism, etc. Nothing is stopping them! Someone can reduce their consumption of animal products without calling themselves vegan, and then become vegan if/when they’re willing to make the full commitment. We can obviously encourage and celebrate that reduction without pretending it satisfies the end goal.
Thus, I don’t really see what lowering the standard accomplishes. The people who want an intermediary path already have several, while existing vegans would be given explicit social permission to eat more animal products and gradually weaken their moral convictions. Unless you have strong evidence that the expectation of perfection is preventing a massive number of people from reducing their consumption at all, this doesn't seem like a good sentiment to propagate.
Good stuff! I'd rather have someone go from eating meat every meal to every other meal to someone go from once a week to full vegan, though obviously both would be best. Appreciating the gray nature of Veganism and morality is tricky to do.