A Response to Negative Utilitarianism

Stephen Casper, thestephencasper@gmail.com

Euconoclastic blog series

Something really interesting to me is that utilitarianism can be summed up very well in 4 words as “the greatest happiness principle,” but at the same time, there’s a trove of different flavors, twists, variants, and deviants to it. One of those types (actually more than one) is negative utilitarianism. Different negative utilitarian theories in one way or another place more emphasis on the moral importance of reducing suffering than promoting happiness. I think it’s hard to defend, but a lot of modern utilitarians and Effective Altruists (EAs) believe in it. There’s even an entire EA organization with negative utilitarian motivations focused on suffering-based ethics research.

So let me first frame the issue and second, maybe convince you that plain old utilitarianism is a better belief which can even capture some of the intuitions of negative utilitarians.

If you’d like to learn about all things utilitarian, I recommend first looking at this other post of mine. One important thing to note is that I’ll use suffering to refer to any undesirable feeling and happiness to refer to any desirable one.


First, let’s consider a non-exhaustive, symmetric list of 6 types of utilitarianism.

1. Absolute (Strong) Negative: either only suffering matters or suffering matters infinitely more than happiness (i.e. that we are allowed to care about happiness but only when making judgments about two states with equal suffering)

2. Threshold Negative: there exists some threshold of pain such that, if any suffering surpasses it, no amount of happiness can outweigh it.

3. Asymmetric (Weak) Negative: suffering and happiness are always commensurable, but suffering should be given more weight.

4. Asymmetric (Weak) Positive: suffering and happiness are always commensurable, but happiness should be given more weight.

5. Threshold Positive: there exists some threshold of happiness such that, if any happiness surpasses it, no amount of suffering can outweigh it.

6. Absolute (Strong) Positive: either only happiness matters or happiness matters infinitely more than suffering (i.e. that we are allowed to care about suffering but only when making judgments about two states with equal happiness)

It’s worth noting that 3 is common, 2 and 4 are uncommon, and 1, 5, and 6 are virtually unheard of.

Also, feel free to learn more about negative utilitarianism from Brian Tomasik here and here, and look at Toby Ord’s response to negative utilitarianism here.

Rejecting the Absolute Views (1 and 6)

The trouble with these views is easy to see. Suppose for concreteness that we adopt an absolute negative utilitarian view (the argument against the absolute positive view is perfectly symmetric). Imagine a very happy world where everybody had an extremely high quality of life and nobody ever had an unpleasant feeling. But then imagine that someone stubbed their toe before immediately shrugging it off and going back to feeling fantastic. If we were absolute negative utilitarians, we’d consider this world to be net-negative.

It’s crucial to note that this isn’t necessarily a problem — no contradictions or absurdities to be found here. It’s only a problem in light of a value system that already assumes the falsehood of the absolute negative view. But my point is that (almost?) everyone has such values. If the axioms we choose to build our moral system on stipulate that suffering is infinitely more important than happiness, then the only argument to be made is to point out that axioms disagree. But for a human being to believe in this would be extremely weird, and probably impossible given our biology. I don’t think anyone exists who thinks that either happiness or suffering is infinitely more important than the other. If any such person existed, I’d be very curious to hear them explain why.

To pound another nail in the coffin, we can observe that (almost?) every person makes daily tradeoffs between happiness and suffering. For example, I might eat junk food which makes me feel good in the short term but be bad for my health in the long term. Or I might go to the gym and strain myself in the short term to feel better and healthier in the long term.

Rejecting the Threshold Views (2 and 5)

Again, I’ll frame this in terms of the negative threshold view, but there’s a perfectly symmetric argument against the positive threshold one. Suppose we adopted the negative threshold view. Then we would believe that there is some level of suffering at which — wait hold on a second — what level of suffering? That’s our first problem. How intense would such a level of suffering be, and what would make it so special? This seems highly arbitrary. But okay, okay. Let’s assume there exists some threshold. We would believe that suffering less intense could be outweighed by enough happiness and suffering more intense could not.

The problem here is that it would cause us to believe that suffering just beyond this threshold, perhaps only even crossing it by a single spike of a neuron, would be infinitely worse than a slightly less bad form of suffering. If we use human preference as a proxy for value, we could actually test how well a theory like this squares with human experience. For example, we could talk to someone who experienced a gradually-increasing and intense level of suffering — maybe someone like Giles Corey. We could ask this person if there was ever a moment when their suffering surpassed a threshold that made their experience suddenly infinitely worse than it was a moment ago. I suspect that Corey would not say this was the case.

Again, we could hold fast to this view by baking the corresponding values into moral axioms, but I don’t think these are values that many people if any at all actually hold.

And consider tradeoffs again. Suppose you were experiencing suffering extremely close to such a threshold. What if you were offered the chance to increase your suffering by a barely-perceptible amount in order to get an immense amount of happiness later in your life? Do you think there could really exist such a threshold that would make you say no?

Still think that some levels of suffering are so great though that they can’t be outweighed by any level of happiness? This other post of mine might be able to convince you otherwise using what I call a continuous commensurability argument. And I don’t mean for this to be a link lazily tossed in here. I recommend you at least skim it if you have the time.

The Asymmetric Views, Properly Understood (3 and 4)

Then we’re left with the moderate views. And you know the drill — I’ll talk about asymmetric negative utilitarianism, but there’s a symmetric argument for the positive version.

Here is an example of asymmetric negative utilitarianism IMPROPERLY understood. We might naively formulate the view as saying that a single unit of happiness is less good than a single unit of suffering is bad.

The trouble with this view (Toby Ord makes a similar point, albeit using a different approach) is that it doesn’t make much sense to put anything on the X axis that would make the curve look this way. It wouldn’t make sense to use units of how desirable an experience is because then, by definition, the curve would have to be linear. We could also try to use some notion of intensity, but that would be arbitrary and not very germane to the task of making moral judgments.

I think that the only sensible kind of view to call “asymmetric negative utilitarianism” is simply the belief that in the real world, most moral subjects’ experiences are net negative.

And this is a good view to have. For example, I think it’s pragmatic to, in many cases, take an approach to doing good in the world that emphasizes the reduction of suffering.

The Collapse

But wait a minute. At that point, what’s the need for the negative label to our brand of utilitarianism? This just makes us regular utilitarians with certain beliefs about how certain subjects suffer. After all, nothing about regular utilitarianism says that happiness and suffering in the real world have to be symmetric.

Okay, but isn’t this just an issue of what we name our philosophy? Couldn’t we just take or leave the label? Maybe someone thinks that happiness and suffering are perfectly commensurable but nonetheless believes that on Earth suffering happens to substantially outweigh happiness and is thereby of higher moral importance. Wouldn’t it be pragmatic for them to just call themselves a negative utilitarian? It’s literally just semantics, right?

Wrong! (In my humble opinion.) I do think that such a person calling themselves a negative utilitarian is bad. And I don’t think I’m splitting hairs either.

A person with asymmetric negative utilitarian beliefs should be at least open to the idea that for some reference classes, such as people who are wealthy, healthy, or self-actualized, that happiness outweighs suffering. Should they call themselves a negative utilitarian about the global reference class but a positive one about some narrower reference classes? They could, but that could be confusing and complicated. Sure, there may exist substantial imbalances of happiness and suffering in different agents, and sure, it’s possible that the suffering on Earth outweighs the happiness. But why would we stick our viewpoints into a restrictive box labeled “negative utilitarianism”?

So I think that the viewpoints behind asymmetric negative utilitarianism should really just collapse into regular utilitarianism, albeit with an asterisk noting that net experience may be negative.

But Isn’t This Just Getting Pedantic?

Wait! There’s one more point I need to make. I think that negative utilitarianism is also a tragedy of emphasis. I think that focusing on suffering — especially if it’s codified into the name we give our moral theory — underemphasizes the moral opportunities of the future.

Think about the last 100 years. How much better has the average human quality of life increased over that short time period? And how much better do you think we could make life for humans and non-humans alike in the medium and long term futures? If we can preserve civilization long enough, we will likely be able to eradicate poverty, achieve peaceful globalization, cure most diseases and mental health issues, genetically modify ourselves to be happier and healthier, engineer our biosphere to reduce suffering, spread out into the cosmos, and maybe — just maybe — figure out how to build or simulate agents that experience happiness on a large scale.

That is, we could take steps to systematically engineer the very world we live in to be more optimized for happiness.

What’s the happiest you’ve ever felt? It was due to micro/nano-grams of material stimulating a very limited region of your 2 kg brain, a biological machine that is presumably not even close to being hedonically optimized.

My point is that there’s a long way to go up.

The trouble with negative utilitarianism is that it focuses too much (and often in a very speculative way) on biological organisms. And I think that’s too shortsighted. If I were to put an asterisk on my personal brand of utilitarianism, if anything it would be one that’s optimistic about the amount of net happiness that the future can hold.

In A Nutshell

So I don’t think that anything under the umbrella of what people call positive/negative utilitarianism is both defensible and properly named. I think “negative” and “positive” utilitarianism are either misguided, misunderstood, or a failure of emphasis.

Euconoclastic (ajd.) \yu̇-ˈkä-nə,-klast-ic\: iconoclastic in a good and virtuous way. Find me at stephencasper.com.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store