Total Utilitarianism: an unethical ethical theory.

It doesn’t even respect its own principles.

Utilitarianism is a family of ethical theories that are based on the concept of Utility. There are several competing definitions of what Utility is, ranging from Hedonic experiences (pleasure and pain) to preferences. I will not attempt to define what Utility is in the present article, and simply use the broadest, most consensual possible definition: Utility is that which is good to have.¹ The advantage of using such a definition is that it bypasses the differences between preference-based theories of Utility (it is what agents want to have), Hedonism (it is what agents find pleasurable), objective-list theories (it is what its proponents personally believe is good for others to have is objectively good to have), etc.

For the rest of the article, you are free to use whichever definition of Utility that you prefer, as long as it is consistent with the above: for instance, you can’t define Utility as “that which is bad to have” and then claim that my article doesn’t apply to that.²

Utilitarianism, as a broad family, roughly aims at applying the following principle: “The greatest Utility for the greatest number”.³

Total Utilitarianism states that this means we should try to maximize the total sum of Utilities across all agents.

The main axioms of the theory can be roughly summarized as:

• Aggregation: the goodness or badness of a population is entirely determined by the Utility of its members.

• Additivity: Utilities can be added linearly, and what matters is the sum total of Utilities.

• Separability: the contribution of each individual to the total doesn’t depend on the level of Utility of others, nor on the size of the population.⁴

• Impartiality: the point of view to be adopted is that of an impartial observer, or as Utilitarians like to call it: “the point of view of the Universe”. Each individual’s Utility counts equally. Keep this sentence in mind, we will get back to it later.

• Universality/Neutrality: adding one new person adds to the goodness of the population, and this doesn’t depend on the average Utility of its members.

In the present article, I will show that Total Utilitarianism violates any reasonable definition of impartiality, explain why it does that, and propose that it be discarded and replaced with a density-based model.

A tale of two cities.

Note: this story uses words such as “happiness” and “suffering”. It wouldn’t have been very pleasant to read if I had used “Utility” and “Negative Utility” instead.

In her short story “The Ones Who Walk Away from Omelas”, Ursula K. Le Guin describes the city of Omelas, where everyone is happy, but their happiness depends on the suffering of one innocent child who is being tortured relentlessly. Some of the citizens of the city leave after learning about the existence of the kid. I imagine they must create a new city, or else they would just die in the wilderness. Let’s suppose that, for some reason, they can’t obtain happiness and social stability without having a similar amount of pain in the new city as in the original Omelas. They have to make a sacrifice.

To solve this problem, they decide that everyone will take a turn in the torture chamber, and then go back to live happily in the city.⁵ At first it requires enormous sacrifice, for the inhabitants of the new city are few and have to stay in the chamber for weeks at a time. However, as the population grows, the duration of the torture becomes shorter. One week each. Then one day. Then a few hours. Then a few minutes, then a few seconds. Then, as the city becomes a sprawling megalopolis, they create portable devices that make the pain as short as a flicker, as short as the blink of an eye. In the end, the duration of the pain is too short to be noticeable, and the inhabitants simply stop thinking about it at all.

As the elders of Omelas witness the growth of the neighboring city, they realize that the kid need not suffer for the citizens to be happy.⁶ They start debating a way to stop the sacrifice and finally free the child from their prison. At first, they think of asking the citizens to take the pain in turns, but they are soon confronted with a difficulty: nobody wants to do it.⁷ After all, there is a reason why they have decided to stay: the suffering of the child doesn’t bother them as much as it does the ones who left. Unable to find a solution, the elders decide to ask the oracle.

The oracle speaks thus: “Suffering is a matter of intensity, and intensity is a function of density.”

At first, the elders are puzzled. So they call the mathematicians and the physicists, the chemists and the neurobiologists, and of course, the great Magi of Value, the ones responsible for maintaining the magical scales on which subjective value is weighed in this world. After months of discussions, they finally come to the conclusion that the suffering experienced by the one child can be magically redistributed on the entire population so that everyone experiences slightly less well-being. Since Omelas is already densely populated, each person would experience the equivalent of having a dust-speck in their eye for a few milliseconds once a year. The Magi cast their spell, no one in the city notices the change, and the child is finally set free.

Total Utilitarianism is blind to injustice.

Total Utilitarianism does not distinguish between these three cases. For the Total Utilitarian, it doesn’t matter whether everyone experiences a very short or very mild, even unnoticeable pain, or one person experiences unbearable pain, as long as the Total Utility remains the same. It doesn’t matter either if all the happiness in the world is experienced by just one person while everyone else is miserable.

In other words, these two graphs look the same to the Total Utilitarian:

I don’t know about you, reader, but they certainly don’t look the same to me!

One last chart for the fun of it:

There are 100 people in this chart, each with a Utility of 1.2. Of course, for the Total Utilitarian, it is still exactly the same chart, there is no difference whatsoever.

I will name this inability to notice differences in distribution: Distributianagnosia.

Now, my claim is very straightforward: Total Utilitarianism’s distributianagnosia violates every reasonable definition of impartiality.

You can already see this in the two-cities story. However, since I have seen Utilitarians claim that inflicting great pain for the sake of many small benefits is morally justifiable (it’s a core tenet of the theory), I would like to provide an example that is even more striking, and should, I hope, make even the most convinced Total Utilitarian frown their brow and scratch their head.

Imagine “Reverse Omelas”, a city where everyone is wallowing in abject misery, except for Ben Thamite, a random guy who is living a life of unfathomable luxury, thus hoarding all the available Utility for himself.⁸ Now Ben is not exceptionally good at converting resources into Utility, he doesn’t have any particular ability to enjoy things more than the others, he just happens, by sheer coincidence, to own all of the wealth and enjoy all of the pleasures that are available to the population of this city alone. An Impartial Observer happens to descend from the sky and starts observing. What does he see? That Ben is enjoying everything there is to enjoy while the other citizens don’t have anything at all to enjoy, and that Ben is not different from the others: he is not a Utility Monster.⁹ The Observer therefore notes that favoring Ben does not even satisfy any kind of efficiency principle. In other words: the distribution of Utility is just skewed in his favor.

Now suppose a Total Utilitarian emerged from the darkest depths ground and said: “The city is fine. All Utility distributions that lead to the same Total Sum are equivalent.” Do you think the Impartial Observer would agree? Of course not, he would retort: “How dare you! This is a clear case of partiality and favoritism, and it is my duty, as an Impartial Observer, to Observe that!”, and he would write a stern report titled “Observations on a case of non-observance of the principles of impartiality.”

Why does it do this?

Why does Total Utilitarianism, a theory supposedly based on impartiality, violate common sense intuitions about what the results of an impartial distribution of Utility should look like, to the point of warranting a stern report?

Well, if you remember how the axiom of impartiality is stated (“Each individual’s Utility counts equally”), you should note one thing: it doesn’t define impartiality in relation to the agents, but in relation to the Utility itself. In other words, what the theory calls “impartiality” does not translate as “you should be impartial towards each agent,” but “each agent’s Utility counts the same, regardless of where it comes from.” Agents are seen as a source of Utility, not as the recipients of Utility. In other words, Utility is a resource to be extracted from people, not a product to be distributed among people. And of course, who cares where a resource came from once it is extracted?

For this reason, Total Utilitarianism is anti-human. It feeds on humans.

What should replace it?

My preferred alternative is that we should use a density-based model to compare different states of the world instead of a pure aggreation function: for any given amount of Total Utility, all things being equal, the worlds in which Utility is distributed more equally are better than the ones in which it is extremely concentrated. This should help recenter the moral patients as the beneficiaries of Utility instead of its producers.

To do this, we could either use some version of the Rawlesian maximin principle, or develop a full-fledged density model. If you have been reading me for a while, you can already guess that I am more favorable to the second option. Developing such a model will be the subject of a future article.

Thanks for reading! As usual, feel free to subscribe!

1. Why do I have the feeling that this definition will be controversial?

2. “That which is bad to have” is usually defined as Negative Utility.

3. Adapted from Bentham’s “the greatest happiness of the greatest number”.

4. If you have followed me for a while, you already know that I reject this axiom.

5. Assume there are no long-term consequences, neither physical nor psychological.

6. In the original story, this is assumed to be impossible, because it is more focused on the conflict between Deontology and Utilitarianism than on comparing different versions of Utilitarianism.

7. I just realized that there is a striking similarity with factory farming here: everyone agrees that it’s very bad, but almost nobody wants to do anything about it. Here are a few charities if you want to do something to help our fellow Earthlings: https://animalcharityevaluators.org/recommended-charities/

8. The example assumes that the Total Utility remains the same regardless of distribution. You can adapt the size of Reverse Omelas’s population in your imagination until it makes intuitive sense to you.

9. Utility Monsters are a separate problem. I will not talk about them in the present article.