People oftentimes reject total utilitarianism the moment they stumble across the repugnant conclusion, all while ‘quality over quantity’ justifies bringing into existence miserable people to uphold average welfare. Prioritizing quality of life over total utility leads you to demand the suffering of individuals rather than simply creating more lives that are still happy! Though average utilitarianism is loaded with sadism, biting the bullet on the sadistic conclusion may be worth considering.
For the sake of simplicity, we’ll say that welfare is measured in happiness servings—every person can have a maximum of 100 happiness servings, needing 50 to have a life worth living.
The sadistic conclusion is the most obvious pitfall of average util, as it justifies actively torturing and harming people; being considered much worse than the repugnant conclusion—if you have a population containing 1 trillion people, all of whom have 95 happiness servings and could either press:
A button that adds 100 billion people with 75 happiness servings (quite happy people!)
A button that adds 100 people who live in a state of torture, each having 10 happiness servings
The population starts with an average utility of 95 happiness servings. Pressing button 1 yields an average of: (1 trillion * 95 + 100 billion * 75)/1,100 billion, or 93.18 happiness servings per person. Pressing button 2 yields an average of: (1 trillion * 95 + 100 * 10)/(1 trillion + 100), or 94.99999… happiness servings per person. It almost seems as if this theory morally advocates for creating terrible lives! Doing some elementary math, we approach the sadistic conclusion; such an inference is why most people choose to bite the bullet on the repugnant conclusion.
Accepting the sadistic conclusion over the repugnant conclusion entails preferring to materialize a ‘victim’ or a small underclass of people with highly diminished utility over materializing a larger underclass of people with less diminished returns—as long as creating the ‘victim’ lowers the average less than the latter option. Oftentimes, it’s easier to bring into existence one miserable person who can have his/her welfare easily raised to the standards of the societal average as opposed to a larger sum of people constituting a welfare underclass.
The 1973 story, “Those Who Walk Away from Omelas,” depicts a hypothetical society where every resident’s utopian level of happiness (100 servings, if you will) is contingent on a single child’s suffering—a society that’s a negligible amount away from a complete utopia. It’s the literal embodiment of the sadistic conclusion, though in this case, the victim cannot be alleviated from a state of suffering without destroying the optimized happiness of everybody else; a real society would still have the option of helping out the victim.
Pressing a button that would turn every resident on Earth, minus 1, into a resident of a utopian paradise at the cost of the 1 outlier suffering for life is a hugely positive-sum trade, but also a sadistic one that creates a permanent victim who does not have a life worth living. The same cannot be said for the society we live in—negative-sum trades where little to no benefit is incurred at the expense of obscene amounts of suffering are commonplace, while many individuals still view Omelas as morally abhorrent simply due to the juxtaposition between an entire society’s welfare and one child’s suffering.
Society right now is the best it’s been since we humans invented the notion of society, but it’s not nearly as ‘efficient’ or as ideal as the hypothetical Utopia of Omelas, which is itself not as ideal as a total utopia. Let’s compare Omelas to another hypothetical, utilitarian society that faces the repugnant conclusion instead of the sadistic one; we’ll refer to it as Salemo.
Salemo is a society with an absurd amount of people, a behemoth in terms of total utility, with the catch being that every single person within it has 50 happiness servings—the bare minimum that’s needed to make a life worth living. Would this society be better or worse than its sadistic counterpart? It seems evident that Salemo is not a good place to live in; if Earth’s inhabitants have an average of 70 happiness servings currently (weighed down by immense suffering in underdeveloped nations), it would mean that your life would almost certainly be a lot more miserable if you were to live in Salemo.
Even the most miserable people currently alive on Earth have lives worth living, albeit with a small margin of happiness between their life and a life barely worth living. Just imagine what would happen if you made their life even worse, then imagine quadrillions of people all living life like them.
Which society would you rather live in?
Salemo, where you have a 100% chance of living a life barely worth living
Omelas, where you have a 99.9999% chance of living a perfect, utopian life (this is even a conservative estimate; a large city having only 1 million people is unlikely)
If you’re rational, you’d choose to live in Omelas. If I told you to either press a red button that has a 100% chance of making your life much worse, versus a blue button that has a 999,999/1,000,000 chance of making your life the best it could possibly be, albeit with a 1/1,000,000 chance of making your life total suffering, pressing the blue one is a must.
Besides this, there are many other hypotheticals that point towards embracing the sadistic conclusion:
You can either genetically engineer 1 billion people with 50 happiness servings each or 1 million people with 95 happiness servings each. It’s intuitive to bring into existence the 1 million people, rather than inflating the world with people with lives barely worth living.
Your grandpa is living a joyous, happy life with 90 happiness servings. However, he is now 90 years old and will pass away soon, but we have a machine that can extend his life by 20 years; unfortunately, those years will be plagued by dementia and illness, making it so he barely has any happiness servings in his last 2 decades. His life will still be worth living nonetheless.
Should you artificially inflate your grandpa’s life by 20 years? Doing so would be quite immoral, though it still increases utility under total utilitarianism. We also know that we don’t have an obligation to produce as many babies as possible for the sake of achieving maximum utility. But before committing to any of these hypotheticals, consider refutations to them—do they counteract the justification given to sadism by the hypotheticals?
The Omelas hypothetical only proves that the interests of the many outweigh the interests of an individual, something that all versions of utilitarianism should accept. Preferring Omelas over Salemo only means that there exist some cases where an extremely weak instance of sadism is preferable to an extremely potent instance of repugnance, not that biting the bullet on the repugnant conclusion is now less rational than doing so in the case of the sadistic conclusion.
Genetically engineering 1 million people with 90 happiness servings sounds intuitively better than genetically engineering 1 billion people to have 50 happiness servings (barely worth living), but that’s because of resource scarcity, not the validity of average utilitarianism. If we had the resources available to raise the welfare and accommodate those 1 billion people, it’s evident that we should bring those 1 billion people into existence instead of only 1 million people.
Your grandpa could’ve also had a relatively happy 20-year life extension despite bringing his average down—obviously leading you to choose extending his life, so long as we raise the average welfare of the extension.
The repugnant conclusion seems unacceptable, as it implies that you have to fill the world with as many people as possible to maximize the utility curve. The sadistic conclusion seems unacceptable, as it implies that you should create miserable people so long as creating happy people doesn’t decrease the average as much!
To commit to utilitarianism here, you’re going to have to pick your poison.
Nice post! I have two things I'd want to object to though:
First, when you compare Salemo (many many many people with lives barely worth living) to Omelas (fewer very good lives, and a single bad life), you lay out the options of either choosing to live in:
1) Salemo, where you have a 100% chance of living a life barely worth living
2) Omelas, where you have a 99.9999% chance of living a perfect, utopian life (this is even a conservative estimate; a large city having only 1 million people is unlikely)
But this framing is wrong! This assumes that you have a 100% chance of existing at all in both cases. Total and average utilitarianism only differ in cases where the number of people differ, so it's a big nono in this type of comparison to assume that you'll be sure that you exist! So the options should be a small chance of a very good life vs a much much much larger chance at a life barely worth living. I think that makes the comparison turn the other way.
Second, you describe adding 20 years of life to your grandpa, which "will be plagued by dementia and illness, making it so he barely has any happiness servings in his last 2 decades." In the last line you stipulate that the life is barely worth living. But I don't think the years you describe are necessarily ones worth living!
I think it's very hard to asses how good a life has to be worth living. You can always stipulate that a life in a thought experiment is worth living, but that doesn't help. I can say that Bob lives in fire and brimstone, being burned and poked day in and day out--but I stipulate that his life is worth living. Okay sure, but obviously I'm gonna judge creating Bob as a negative thing, since the life I'm actually imagining is not one worth living, despite my stipulation.
I do find sadistic conclusion more palpable than repugnant one.
But it seems to me that the worsts offenses of both can be moderated if we embrace some kind of "complex utilitarianism" that depends on both total and average utilities, weighted by certain coefficients and combined together.