I wonder what the best “yes but” responses to this thesis are? What about the idea for example that creating more income equality creates more happiness by flattening the comparison curve? Isn’t that a way of bridging your critique with a policy that still concerns itself with happiness? Same with addressing all tractable forms of suffering to which people do not hedonically adapt to like depression, anxiety and chronic pain or living in noisy, substandard housing with bad air quality and without enough money for a nutritious diet and clothes that allow one to look typical.
And just on the Nozick’s experience machine. I’ve always found his formulation unpersuasive as evidence that people don’t really want an easier life of more contentment. Entertain for example the alternative formulation where you stipulate one’s present life as being in the experience machine and the reality choice as being in a Soviet gulag for life with only rotten bread to eat, a bucket for a toilet and semi-regular beatings. Also, look at how many people take anti-depressants. That’s a form of the experience machine (and not even a particularly good one — imagine a much better drug with no side effects and much higher efficacy).
The alternative formulation doesn't prove that the experience machine isn't bad, it only proves that there is some level of real-life misery that is even worse than an experience machine. A better alternative formulation would be that you are in an experience machine, and if you left your life would be worse, but still worth living. You'd be poorer, sicker, have worse friends and family, but still enjoy life.
In that case some people at first might want to stay in the machine, but I think that's because they haven't thought it through. In the machine, your friends and loved ones are an illusion. When you kiss your spouse every morning, no one felt the touch of your lips. When you made your baby laugh, nothing really laughed, it was an illusion the machine made, your baby that you love so much never existed. Everyone you ever knew and loved isn't just dead, they never were. Every accomplishment you worked hard on never happened, the machine faked it.
I think when most people picture the alternative formulation and say they wouldn't leave the machine, they are subconsciously analogizing the machine to multi-player videogames. However, those games are not like Nozick's experience machine. The people you interact with in them are real, you can form real friendships and make a difference in real people's lives. Yout accomplishments are real too, you have to put in real effort to beat the game, it is possible to lose. That is not what Nozick's machine was like.
If you say that people might not prefer a classic Nozick experience machine, but might prefer one more like a multi-player game, that might be true, but at that point you have erased the most significant ways the machine differs from reality. If the machine is basically like real life, but safer and more fun, then a modern city is basically an experience machine compared to a medieval village.
I think what it shows is that you can get different results in terms of how people respond to the question depending on how in detail you set up the experiment. The question of whether entering it is bad from a third person philosophical point of view is a different one (and even that one would probably give rise to much back and forth).
Also I’m not sure about the quality of realness. How do we know for sure this life is real? To know for sure you need something to compare it to but in this life and in the experience machine you won’t have that. It’ll all seem real.
Finally if the point of the experiment is to show that there are higher pleasures like contentment with the person one is and the life one is leading I would just concede that right off the bat.
In terms of reducing inequality - I agree that it's a good thing to aim for, but I'm sceptical that it would increase happiness. In the current (unequal) status quo, you're poor you compare yourself to other poor people and you're averagely happy; if you're rich you compare yourself to other rich people and you're still averagely happy. There's not much correlation between happiness and wealth! Does Elon Musk strike you as a particularly happy person, for example? Au contraire, to me he seems bitter, paranoid, and mentally unwell.
Basically the takeaway I took from this piece is "basically the hedonic treadmill means that whatever society does won't move the needle on human happiness anyway, so maximising happiness is a useless directive to follow."
One question I’d have is who, empirically, people compare themselves too. I would guess far more with the people they spend the most time with and around in their daily lives.
Exactly. And society is very efficient at filtering people by wealth. If you're rich you hang out with other rich people; ditto for poor people; and also for those on middle incomes.
"Men who lose their jobs seem to experience a durable negative impact on their well-being."
I was, for a short time, voluntarily out of the labor force. I did experience poor well-being and in my case, I believe it was because of 2 reasons.
One as you mentioned was that my expectations were not my condition, but what my peers were doing.
Second was that leaving my job reduced my status that I earlier had in my hostel. I was living in a hostel where status was determined by how good of a company you were working for.
Hostels, and in general any place which aggregates high agency people is like "a bigger pond' that you mentioned in an earlier post. It makes you ambitious and work hard, but takes a toll on your well-being.
Is there a way to create social places where you're with smart people that makes you ambitious and learn many things while not taking a toll on your well-being because of comparisons?
Hi Ram, thanks a lot for your testimony. It is very much in line with the explanations presented in the post. In regard to your question. I guess it would be possible if you could do so without changing your expectations to compare yourself to them. That could be possible if, for instance, this group is out of bounds for you to join as a member. You could imagine, for example, a group of Buddhist monks. An alternative could be a group of robots that are very smart and can teach you many things. We may be close to that soon!
What do you think of Sen and Naussbaum's Capabilities approach?
"Well-being" and "satisfaction" tend to measure something other than hedonia, though measuring any such thing is a scattershot scatterplot onto any number of things.
In my mind, the "measure" of a society is that it turns good intentions into good outcomes. The idea wouldn't be to actually measure it and use that number in some calculation, but simply to recognize that it is both a resource to tap and a communal responsibility as much as a social mandate of representatives. Bridging that gap is also related to GDP, satisfaction and capabilities without overcommiting to any one variable or angle.
It's easier to understand than capabilities per se, but similarly hard to measure. However, I'm not sure that's a weakness. Negotiating and debating good intentions and outcomes are part of the process of enacting them and keeping them informed on how to actually bridge the gap.
Hi William, thanks for your feedback. Finding a political philosophy that is consistent and can help guide public policy is difficult. Utilitarianism is intuitive. While I criticise it here, in many cases a utilitarian approach will give conclusions similar to other approaches. Sen's capability approach is an attempt at overcoming utilitarianism's shortcomings. I don't think it works: what are capabilities exactly and how do we measure them? But the intuition makes sense: it aims to give people "real freedom" (to take the book title of Philippe van Parijs). As I indicate in the final footnote, my perspective is contractualist, because I think this is how our psychology works.
A Rawlsian! Excellent. Have you read Sen's Idea of Justice? It primarily seems addressed directly to Rawls.
As for my perspective, the contractualist approach essentially emerges from Sen's capabilities approach as a base level and the pragmatic means of connecting good intentions and good outcomes at a higher level. It would be totally fair to criticize my approach as imprecise if I meant it as a prescription for some top-down policy.
I would, for example, agree that a contractualist approach fits well with and even stabilizes our psychology. However, I think the common, Dawkinsian Tit-for-tat game theoretic accounts misstell the origins of that fit. I won't belabor that point here though.
Instead, I would say the reason to see contractualism as a natural connector is exactly because of the problem you identified with the capabilities approach. What is it in general? The same critique doubly applies to my approach. What are good intentions and good outcomes in general?
But consider that the intuitive-yet-vague landscape is nevertheless tapped when negotiating a contract, and to the degree that capabilities are written larger or smaller, whether good intentions are leveraged more or less, whether the outcomes are narrow or spread, the contract emerges from that negotiation. And the ability to negotiate a contract would be one of Sen's many capabilities as his idea of freedom does stem from economic and social development.
Yes I read Sen and I like it, but (like many theories in political philosophy) it does not hold well when asking foundational questions. My perspective, to be precise, follows Binmore's naturalistic take on contractualism in his book "Natural Justice". I will write about it here at some point. There is a lot to say.
I may respond to this overall idea in a post on the other side of the holidays, but it seems to be a rather grounded means of identifying equilibria. What I wonder, from afar, is how well it is conducive to insights once given specifics of a certain situation. But that is on me to find out with some more research!
I couldn't access the article, but I saw it mentions Mackie. I will be posting an extension of Mackie's concept of INUS conditions to include AynT conditions (Arbitrary yet non-Trivial). This sounds like it lands solidly in the Nash equilibrium space. However, it is more about negotiations under forward thinking uncertainty than what is available as INUS conditions in hindsight.
For a bird's eye view, we might think of how the actions of a person are increasingly predictable with increases in hunger. However, at some set of INUS conditions free from all such "needs" which explain decisions in retrospect, what might be projected forth is poorly approximated by past-facing inquiry.
In the meantime, if you find yourself so inclined, this post contains part of my argument for why statistics is primarily exploratory, a use which is just as important as predictive and explanatory power, but often results in settling for predictive power only. For example, I do not think game theory explains much, despite its undeniable power for navigation.
For example, imagine if people who are misunderstood become more predictable as a way to attempt to be understood. This would be consistent with Nashville equilibrium, but it would be an explanation of nash equilibrium more than vice versa. This is the part that gets laundered.
Anyway, Happy Holidays, and thank you for the friendly engagement!
Thanks! I am not sure I followed all your arguments. I am bullish on game theory as being useful to explain real behaviour. I believe it is relevant for people who are used to interacting repeatedly in the same type of strategic environment (see this post for instance: https://www.optimallyirrational.com/p/anyone-for-tennis-and-game-theory) and for organisms that have been selected by evolution for interacting with other organisms in specific strategic settings. I hope to further demonstrate this in future posts!
This fantastic article makes for a perfect example. Consider the following:
Your study highlights a clearly non-trivial overlap between distributions of shots and energy expenditure at the top levels of tennis. But what does it mean in this specific case to have "explained" behavior? You have explored and impressively predicted "statistical behavior." I assume this is thought to reinforce the explanatory link that evolution by natural selection converges towards various equilibria, and that among various equilibria, powerful complex systems may converge toward nash equilibria.
But this also raises the question of what can be done with that information if methodical study reveals instinctual behavior converges on complex equilibria. How would one then claim that one should ever be methodical and empirical as opposed to intuitive, instinctual and otherwise non-obstructive? After all, the tennis players achieved nash equilibrium by focusing on the task, not on the maths.
Note that I am not claiming we should reject the empirical for the intuitive, but instead calling attention to what is thought to be "explained" by successful statistical prediction. A statistical link to evolutionary theory can be both compelling and beg the very question being asked. If natural selection explains the simpler set of equilibria, should we then assume it is also necessary and sufficient to explain all other complex equilibria thereafter? Is the "else extinct" clause of genetic competition actually explanatory, or is it jointly sufficient for intellectual equilibria among modern scholars? I suspect the latter, but the burden of proof is on me for such a claim. I will of course have to argue that elsewhere and stop blowing up your comment section.
If governments and/or individuals should not optimise for happiness, and everybody habituates to higher and lower levels of success, what would be more useful things for countries/individuals to strive for?
That's a very good question. I will definitely need to write a post on this. One initial answer is that, while happiness is not a quantity to be maximised, people generally enjoy improvements over the status quo, so seeking improvements makes sense. Another answer is that a government's primary role might be to find ways to organise the rules for allocating benefits and costs in society, ensuring that these decisions are consensual and conflicts are addressed through peaceful and democratic means.
What is interesting is that from what I can see, to value continual improving your circumstances as an individual, and setting up "fair" rules (rules for allocating costs and benefits) that a majority of the population can agree on, and ensuring safety for the population, seems to me like values that can be derived from primary rewards, and some of the very things that happiness makes us strive for.
I would love to hear what you think about this, and I look forward to reading your article on the topic.
There are variants of utilitarianism that try to account for other values besides happiness. Preference utilitarianism is probably the most famous example. Derek Parfit's "Success Theory" is the one that seems most sensible and complete to me, although I am sure there are things that it is missing.
I think the value of utilitarianism comes not from its definition of happiness as the only good, which is wrong headed. Rather, it is its view that what is good in life is something people should increase, and that actions that increase it more or for more people are proportionately more good. I think people can endorse this general ethical idea without endorsing utilitarianism.
Whatever you define as good, in normal situations all the variations of utilitarianism agree. Making sure more people are healthy, safe, economically secure, etc, maximizes pretty much any plausible formulation of the good.
You’re right that utilitarianism’s strength lies in its emphasis on increasing what is good in life for more people. This general ethical intuition resonates with many frameworks, not just utilitarianism. However, preference utilitarianism, like classical utilitarianism, faces significant challenges in practice. Without a universal metric for welfare, it becomes difficult to handle trade-offs between subjective experiences or to compare and aggregate across individuals. While these issues are abstract, they have real implications for policy-making.
I agree that in normal situations, different formulations of utilitarianism (and indeed many ethical theories) converge on similar goals—health, safety, and economic security. However, the edge cases where these formulations diverge highlight important limitations that are worth exploring. I’ll touch on contributions from theorists like Harsanyi and Binmore in later posts, as they do provide some interesting answers to these questions. I'll likely discuss Parfit’s “Success Theory” then.
If our brain just uses happiness for whatever it optimizes, than why do we use the brain to optimize for happiness? How do you reconcile this seeming duality?
If they are efficiently designed, happy feelings should reflect differences in the underlying fitness values of the options we face. Our brain seems to chase happiness because it is a guide to good decisions when picking an option.
What if we replace the word 'happiness' with the word 'contentment'? Then rather than having to continually exceed expectations one would only have to recognize it could be worse.
Amazing post. To your last point, Utilitarianism fails not just because evolution doesn’t care about happiness, but also because too much of it is anti-evolutionary. Dissatisfaction is an essential tool of progress. A satisfied society will always be outcompeted by a frustrated one since the former will have no reason to make scientific advances.
Chasing happiness itself is a bit like chasing grades -at all costs- in school, instead of knowledge. At some point, you start hacking the system and get good grades without learning. This would be equivalent to Huxley’s “chemically induced euphoria”. If everyone was in that state, society would crumble.
I coincidentally raised many of the same points in my latest post about Materialism. I think that focusing too much on KPIs that we can measure and improve always leads to integrating this progress as a new baseline which will neutralize the improvement. To me this happens as well on a broad societal level and not just on the personal, psychological one (that is, the hedonic treadmill).
“At some point, you start hacking the system” You are absolutely right. Any incentive system faces this challenge. Our hedonic system is designed to try to prevent this to happen. We adapt to whatever we have and become frustrated or bored, which nudges us to avoid complacency and continue aiming higher.
Excellent and thought provoking essay as always. I wondered why you didn’t consider religion though? I mean it’s obviously a collective coping mechanism for uncertainty? Religion strips people of agency & free will. We look at this now & see it as superstitious nonsense but of course it’s an incredibly powerful way to reduce anxiety / promote happiness. When you are repeatedly taught heuristics: God wills it, life is suffering, be grateful for what you have etc it is much easier to habituate than when the heuristic is: you control your own future, you are responsible for the path your life takes, anyone can win with enough effort. It must have been easier to accept relative discrepancies in wealth when you felt you had no control. But capitalism & science says: you *will* it. You *control* it. So much of modern mental health is really about reinventing the religious premise of: deal with it, accept it, it’s not your fault, look at the beautiful trees & feel grateful. It is why Buddhist teaching or self help versions of it are so popular in the West. They say, stop planning, stop wanting, stop attachments, stay in the present moment where you’re safe. Anyway, to your point about gov policy, I couldn’t agree more. But there is surely a lot to be said for crafting better collective coping strategies than state ordained medical mental health interventions. Far from reducing anxieties, these have apparently reified it into one of the big oppressors of our time. While it may be true that targetting ‘happiness’ per se is pointless, it is still the case that the loss of religion & collective / community based coping has undermined Western populations’ ability to deal with uncertainty - something that govs may wish to address?
Hi Claire, thanks for the feedback. The idea that religion is a "collective/community-based coping" mechanism is very intuitive, particularly in countries with a tradition of a "moralising god." However, the wide diversity of religious beliefs across societies and cultures suggests that religion is not necessarily such a mechanism. A great book on the topic is Boyer's "Religion Explained".
Otherwise, I think there are good alternatives to the utilitarian approach, particularly those following the contractualist approach. I plan to write a series of posts on this in the near future.
Completely agree with your critical view on utilitarianism as a basis for policy. However, I don't think that Rawls' alternative works. Some sort of contractarian approach might do the job, but not his. Have you read Gauthier's Morals by Agreement? There is also Narveson's The Libertarian Idea that takes a similar approach.
Hi Max, yes I think Gauthier's perspective is more relevant to understanding how we resolve social choices. Binmore's "Natural Justice" is similarly informed by game theory and bargaining theory with an added evolutionary perspective. I will discuss it in later posts.
Of possible interest (and with apologies for the mischievous picture of utilitarians arguing). cc @Bentham's Bulldog @Richard Y Chappell @Richard Pettigrew @Lance S. Bush
“Unfortunately, utilitarianism is bound to fail. It is founded on misguided intuitions about what happiness is...This misunderstanding of happiness is a major flaw in the foundations of this philosophical approach, one that makes it unsuited to be what guides public policy.”
I have two objections to this.
First, I have an objection about the nature of happiness as understood by utilitarians. The criticism above rests on a misunderstanding of how sophisticated utilitarians conceptualise happiness. Utilitarians are not committed to a naive conception of happiness. Indeed, this is one of the reasons that contemporary utilitarians talk more in terms of overall well-being or welfare, rather than happiness. The criticism that utilitarians are seeking the repetition of an inherently transient mental state therefore misses the mark as a criticism.
Second, I have an objection about the timeframe over which utilitarians seek to maximise well-being. Utilitarianism does not commit you to the maximisation of well-being by doing, now, whatever would increase well-being most, now. This has long been understood as a flawed way of maximising well-being over the timescale of a human life. Utilitarianism is consistent with forms of suffering that are conducive to increases in overall well-being, like training to become a soldier or undergoing a PhD. Longtermist utilitarians even go well beyond the timescale of a single human life and seek to maximise well-being over the very, very long term.
Hi Michael, thanks for your comment. Here are answers:
“First, I have an objection about the nature of happiness as understood by utilitarians. The criticism above rests on a misunderstanding of how sophisticated utilitarians conceptualise happiness. Utilitarians are not committed to a naive conception of happiness. Indeed, this is one of the reasons that contemporary utilitarians talk more in terms of overall well-being or welfare, rather than happiness. The criticism that utilitarians are seeking the repetition of an inherently transient mental state therefore misses the mark as a criticism.”
- My criticism is mostly targeted at classical utilitarianism, which is the original and most intuitive version of the approach. Modern utilitarians, like Peter Singer, often adopt a “preference” framework, partly to address the conceptual problems ‘of the classical approach. I pre-empted this point in a footnote in the post. This conceptual shift introduces deeper issues. Once happiness is no longer treated as a measurable and aggregable quantity, the objective function of utilitarianism becomes unclear. What exactly is being maximised, and how should trade-offs be handled? This creates a more complex, arguably less coherent framework.
“Second, I have an objection about the timeframe over which utilitarians seek to maximise well-being. Utilitarianism does not commit you to the maximisation of well-being by doing, now, whatever would increase well-being most, now. This has long been understood as a flawed way of maximising well-being over the timescale of a human life. Utilitarianism is consistent with forms of suffering that are conducive to increases in overall well-being, like training to become a soldier or undergoing a PhD. Longtermist utilitarians even go well beyond the timescale of a single human life and seek to maximise well-being over the very, very long term.”
- That is a fair point, but this approach requires principles for adjudicating trade-offs between short-term and long-term well-being. If classical utilitarianism is set aside, it becomes unclear what principles should replace the maximisation of happiness as an experienced hedonic feeling.
I will add some further thoughts, but will keep them relatively short to avoid a prolonged exchange (on the assumption this is not desirable).
Sticking with the language of well-being, rather than happiness, it is quite clear that this can be measured and aggregated to some extent. Well-being comes in degrees, some people have higher levels than others, and at least some of these degrees of difference can be readily recognised. It is still well-being that is being maximised overall. The phenomenon of habituation and relative comparison makes our understanding of well-being more sophisticated and tells us that we cannot simply add more units of, say, wealth for more corresponding units of well-being. Problem cases and precise formulations are a problem for academics, but policy makers could settle on some generally agreed upon metrics with broad consensus.
Perhaps a more interesting and controversial question is how commensurable differing degrees of well-being are and how these can be precisely formulated (for example, should they be understood as one unit along one spectrum, or multiple units along multiple spectrums? Is this the same for all species, or is it species-relative?)
Where you say the framework becomes less coherent, I am inclined to say that it simply becomes more sophisticated. However, I do suspect that the more sophisticated the framework becomes the less clearly action-guiding it becomes. This is a problem for a moral theory if it doesn’t just want to offer a criterion of rightness, but also wants to offer moral agents a decision-making procedure. This touches directly on your second comment, which I agree is a key area for discussion. Even if we accept the maximisation of well-being as the criterion of rightness, how does the sophisticated utilitarian make decisions to adjudicate trade-offs?
You say, "it is quite clear that this can be measured and aggregated to some extent." I think this is actually a very complex question. Whether it is measurable within an individual has been at the centre of debates in economics and psychology for decades. See, for instance, this book for a historical perspective:
My take is that modern neuroscience and evolutionary reasoning suggest that happiness, well-being, or utility does not behave like a quantity that can be measured.
Assuming it could be measured, there would still be the problem of aggregation: how to add units of well-being across different individuals. Even for utilitarians, this is a major issue.
Utilitarians who opt for a "preference" approach are, in my view, akin to economists adopting an ordinalist position about utility (utility is just representing preferences for some things over other things). This makes aggregation conceptually impossible because there is no underlying quantity to measure and aggregate behind the preferences.
For a quick glimpse at my more fundamental take on utilitarianism:
I think utilitarianism is intuitive at some level because we can use it in certain social situations. Binmore, in his book "Natural Justice", points out that utilitarianism is feasible in social contracts where there is external enforcement of the rules of the social contract (e.g., children can accept utilitarianism enforced by parental rules). However, at the societal level, there is no external enforcement—there is nothing above society to enforce a social contract. Therefore, a utilitarian contract is always bound to be contested and broken by dissatisfied parties. It cannot be a stable solution. I will write about this at length in future posts.
Thank you for an engaging and generous exchange, Lionel. You’ve given me lots of useful points to consider.
I’m still of the view that overall well-being comes in degrees, that this varies within and between individuals, and that such variation can be understood in terms of other (perhaps first order) qualities, such as being in better or worse health. But I’m also more sympathetic to some of the complexities you have raised.
Many thanks for the book recommendation, too. Most of my current books on ‘utility’ are works of philosophy, but I’d be very interested to add works from economics and psychology. An overview, such as the one you mention, fits the bill perfectly.
I will subscribe to your page and I forward to reading more of your posts in future. Wishing you all the best.
This is a problem with how happiness is measured (with reference to an ordinal scale), not a problem with happiness. As I've said before it's like measuring height by asking children whether they are tall, medium or short. You'll get much the same distribution whether you ask them at age 5 or at age 15. So, you could conclude that even though some children are taller than others, no one gets any taller over time.
The hedonic treadmill is as you well describe, but your characterization of utilitarianism is skewed similarly to how opponents’ characterizations of deontology include rigid and absurd adherence to rules even when a particular situation clearly calls for deviation. Many utilitarians, counter to how you represent them, seem at times overly concerned with future people/consequences (climate change, AI misalignment, etc.) at the expense of current people and happiness.
I’m not a utilitarian or a deontologist or a pragmatist or anything else because I see the obvious strengths all these philosophies offer depending on circumstance. All are insightful, and the danger lies in dogmatically accepting or rejecting any one of them at the expense of all others. This of course leads to extremism, which is the one philosophy I see no strength in.
This is a very timely warning for Indian policy makers. The recent spurt in Direct Benefit Transfers right before State Assembly elections will likely be transformed into a spiral of ever-increasing amounts for the beneficiaries-thanks to habituation.
We need to think of potent measures to safeguard our country's fiscal health.
A measurable quantity has specific properties. Length and volume are examples of measurable quantities. Classical utilitarianism treat happiness as such a thing: a quantity that can be accumulated. Happiness is not something like that. Due to habituation, increases in happiness will tend to fade away.
The answer is in the post. Improving conditions at the bottom end has an effect. Not so much for higher levels. What to do is for another post. But there are alternatives to utilitarianism. The most influential one is Rawls' approach. It is likely supported by more political philosophers than utilitarianism.
The non linearity, yes to some extent, though it is more extreme than typically assumed in the utilitarian literature. In addition, habituation and the illusion that habituation will not happen are major challenges to utilitarianism (recognised as such for quite some time).
Fascinating topic.
I wonder what the best “yes but” responses to this thesis are? What about the idea for example that creating more income equality creates more happiness by flattening the comparison curve? Isn’t that a way of bridging your critique with a policy that still concerns itself with happiness? Same with addressing all tractable forms of suffering to which people do not hedonically adapt to like depression, anxiety and chronic pain or living in noisy, substandard housing with bad air quality and without enough money for a nutritious diet and clothes that allow one to look typical.
And just on the Nozick’s experience machine. I’ve always found his formulation unpersuasive as evidence that people don’t really want an easier life of more contentment. Entertain for example the alternative formulation where you stipulate one’s present life as being in the experience machine and the reality choice as being in a Soviet gulag for life with only rotten bread to eat, a bucket for a toilet and semi-regular beatings. Also, look at how many people take anti-depressants. That’s a form of the experience machine (and not even a particularly good one — imagine a much better drug with no side effects and much higher efficacy).
Just some thoughts.
The alternative formulation doesn't prove that the experience machine isn't bad, it only proves that there is some level of real-life misery that is even worse than an experience machine. A better alternative formulation would be that you are in an experience machine, and if you left your life would be worse, but still worth living. You'd be poorer, sicker, have worse friends and family, but still enjoy life.
In that case some people at first might want to stay in the machine, but I think that's because they haven't thought it through. In the machine, your friends and loved ones are an illusion. When you kiss your spouse every morning, no one felt the touch of your lips. When you made your baby laugh, nothing really laughed, it was an illusion the machine made, your baby that you love so much never existed. Everyone you ever knew and loved isn't just dead, they never were. Every accomplishment you worked hard on never happened, the machine faked it.
I think when most people picture the alternative formulation and say they wouldn't leave the machine, they are subconsciously analogizing the machine to multi-player videogames. However, those games are not like Nozick's experience machine. The people you interact with in them are real, you can form real friendships and make a difference in real people's lives. Yout accomplishments are real too, you have to put in real effort to beat the game, it is possible to lose. That is not what Nozick's machine was like.
If you say that people might not prefer a classic Nozick experience machine, but might prefer one more like a multi-player game, that might be true, but at that point you have erased the most significant ways the machine differs from reality. If the machine is basically like real life, but safer and more fun, then a modern city is basically an experience machine compared to a medieval village.
I think what it shows is that you can get different results in terms of how people respond to the question depending on how in detail you set up the experiment. The question of whether entering it is bad from a third person philosophical point of view is a different one (and even that one would probably give rise to much back and forth).
https://www.psychologytoday.com/us/blog/finding-meaning-in-imperfect-world/201903/pleasure-or-reality-the-experience-machine-debate
Also I’m not sure about the quality of realness. How do we know for sure this life is real? To know for sure you need something to compare it to but in this life and in the experience machine you won’t have that. It’ll all seem real.
Finally if the point of the experiment is to show that there are higher pleasures like contentment with the person one is and the life one is leading I would just concede that right off the bat.
In terms of reducing inequality - I agree that it's a good thing to aim for, but I'm sceptical that it would increase happiness. In the current (unequal) status quo, you're poor you compare yourself to other poor people and you're averagely happy; if you're rich you compare yourself to other rich people and you're still averagely happy. There's not much correlation between happiness and wealth! Does Elon Musk strike you as a particularly happy person, for example? Au contraire, to me he seems bitter, paranoid, and mentally unwell.
Basically the takeaway I took from this piece is "basically the hedonic treadmill means that whatever society does won't move the needle on human happiness anyway, so maximising happiness is a useless directive to follow."
One question I’d have is who, empirically, people compare themselves too. I would guess far more with the people they spend the most time with and around in their daily lives.
Exactly. And society is very efficient at filtering people by wealth. If you're rich you hang out with other rich people; ditto for poor people; and also for those on middle incomes.
"Men who lose their jobs seem to experience a durable negative impact on their well-being."
I was, for a short time, voluntarily out of the labor force. I did experience poor well-being and in my case, I believe it was because of 2 reasons.
One as you mentioned was that my expectations were not my condition, but what my peers were doing.
Second was that leaving my job reduced my status that I earlier had in my hostel. I was living in a hostel where status was determined by how good of a company you were working for.
Hostels, and in general any place which aggregates high agency people is like "a bigger pond' that you mentioned in an earlier post. It makes you ambitious and work hard, but takes a toll on your well-being.
Is there a way to create social places where you're with smart people that makes you ambitious and learn many things while not taking a toll on your well-being because of comparisons?
Great post. Thank you.
Hi Ram, thanks a lot for your testimony. It is very much in line with the explanations presented in the post. In regard to your question. I guess it would be possible if you could do so without changing your expectations to compare yourself to them. That could be possible if, for instance, this group is out of bounds for you to join as a member. You could imagine, for example, a group of Buddhist monks. An alternative could be a group of robots that are very smart and can teach you many things. We may be close to that soon!
What do you think of Sen and Naussbaum's Capabilities approach?
"Well-being" and "satisfaction" tend to measure something other than hedonia, though measuring any such thing is a scattershot scatterplot onto any number of things.
In my mind, the "measure" of a society is that it turns good intentions into good outcomes. The idea wouldn't be to actually measure it and use that number in some calculation, but simply to recognize that it is both a resource to tap and a communal responsibility as much as a social mandate of representatives. Bridging that gap is also related to GDP, satisfaction and capabilities without overcommiting to any one variable or angle.
It's easier to understand than capabilities per se, but similarly hard to measure. However, I'm not sure that's a weakness. Negotiating and debating good intentions and outcomes are part of the process of enacting them and keeping them informed on how to actually bridge the gap.
Hi William, thanks for your feedback. Finding a political philosophy that is consistent and can help guide public policy is difficult. Utilitarianism is intuitive. While I criticise it here, in many cases a utilitarian approach will give conclusions similar to other approaches. Sen's capability approach is an attempt at overcoming utilitarianism's shortcomings. I don't think it works: what are capabilities exactly and how do we measure them? But the intuition makes sense: it aims to give people "real freedom" (to take the book title of Philippe van Parijs). As I indicate in the final footnote, my perspective is contractualist, because I think this is how our psychology works.
A Rawlsian! Excellent. Have you read Sen's Idea of Justice? It primarily seems addressed directly to Rawls.
As for my perspective, the contractualist approach essentially emerges from Sen's capabilities approach as a base level and the pragmatic means of connecting good intentions and good outcomes at a higher level. It would be totally fair to criticize my approach as imprecise if I meant it as a prescription for some top-down policy.
I would, for example, agree that a contractualist approach fits well with and even stabilizes our psychology. However, I think the common, Dawkinsian Tit-for-tat game theoretic accounts misstell the origins of that fit. I won't belabor that point here though.
Instead, I would say the reason to see contractualism as a natural connector is exactly because of the problem you identified with the capabilities approach. What is it in general? The same critique doubly applies to my approach. What are good intentions and good outcomes in general?
But consider that the intuitive-yet-vague landscape is nevertheless tapped when negotiating a contract, and to the degree that capabilities are written larger or smaller, whether good intentions are leveraged more or less, whether the outcomes are narrow or spread, the contract emerges from that negotiation. And the ability to negotiate a contract would be one of Sen's many capabilities as his idea of freedom does stem from economic and social development.
Yes I read Sen and I like it, but (like many theories in political philosophy) it does not hold well when asking foundational questions. My perspective, to be precise, follows Binmore's naturalistic take on contractualism in his book "Natural Justice". I will write about it here at some point. There is a lot to say.
I discussed Binmore's view shortly here:
https://www.optimallyirrational.com/p/social-norms-as-rules-of-social-games
For a quick overview of his take: https://www.cambridge.org/core/services/aop-cambridge-core/content/view/9735FE47D3F55920759732BF171A2E52/S1477175609000025a.pdf/justice_as_a_natural_phenomenon.pdf
Much appreciated!
I may respond to this overall idea in a post on the other side of the holidays, but it seems to be a rather grounded means of identifying equilibria. What I wonder, from afar, is how well it is conducive to insights once given specifics of a certain situation. But that is on me to find out with some more research!
I couldn't access the article, but I saw it mentions Mackie. I will be posting an extension of Mackie's concept of INUS conditions to include AynT conditions (Arbitrary yet non-Trivial). This sounds like it lands solidly in the Nash equilibrium space. However, it is more about negotiations under forward thinking uncertainty than what is available as INUS conditions in hindsight.
For a bird's eye view, we might think of how the actions of a person are increasingly predictable with increases in hunger. However, at some set of INUS conditions free from all such "needs" which explain decisions in retrospect, what might be projected forth is poorly approximated by past-facing inquiry.
In the meantime, if you find yourself so inclined, this post contains part of my argument for why statistics is primarily exploratory, a use which is just as important as predictive and explanatory power, but often results in settling for predictive power only. For example, I do not think game theory explains much, despite its undeniable power for navigation.
https://open.substack.com/pub/williamofhammock/p/vindicating-denis-noble?utm_source=share&utm_medium=android&r=3nwud0
For example, imagine if people who are misunderstood become more predictable as a way to attempt to be understood. This would be consistent with Nashville equilibrium, but it would be an explanation of nash equilibrium more than vice versa. This is the part that gets laundered.
Anyway, Happy Holidays, and thank you for the friendly engagement!
Thanks! I am not sure I followed all your arguments. I am bullish on game theory as being useful to explain real behaviour. I believe it is relevant for people who are used to interacting repeatedly in the same type of strategic environment (see this post for instance: https://www.optimallyirrational.com/p/anyone-for-tennis-and-game-theory) and for organisms that have been selected by evolution for interacting with other organisms in specific strategic settings. I hope to further demonstrate this in future posts!
This fantastic article makes for a perfect example. Consider the following:
Your study highlights a clearly non-trivial overlap between distributions of shots and energy expenditure at the top levels of tennis. But what does it mean in this specific case to have "explained" behavior? You have explored and impressively predicted "statistical behavior." I assume this is thought to reinforce the explanatory link that evolution by natural selection converges towards various equilibria, and that among various equilibria, powerful complex systems may converge toward nash equilibria.
But this also raises the question of what can be done with that information if methodical study reveals instinctual behavior converges on complex equilibria. How would one then claim that one should ever be methodical and empirical as opposed to intuitive, instinctual and otherwise non-obstructive? After all, the tennis players achieved nash equilibrium by focusing on the task, not on the maths.
Note that I am not claiming we should reject the empirical for the intuitive, but instead calling attention to what is thought to be "explained" by successful statistical prediction. A statistical link to evolutionary theory can be both compelling and beg the very question being asked. If natural selection explains the simpler set of equilibria, should we then assume it is also necessary and sufficient to explain all other complex equilibria thereafter? Is the "else extinct" clause of genetic competition actually explanatory, or is it jointly sufficient for intellectual equilibria among modern scholars? I suspect the latter, but the burden of proof is on me for such a claim. I will of course have to argue that elsewhere and stop blowing up your comment section.
Cheers!
If governments and/or individuals should not optimise for happiness, and everybody habituates to higher and lower levels of success, what would be more useful things for countries/individuals to strive for?
That's a very good question. I will definitely need to write a post on this. One initial answer is that, while happiness is not a quantity to be maximised, people generally enjoy improvements over the status quo, so seeking improvements makes sense. Another answer is that a government's primary role might be to find ways to organise the rules for allocating benefits and costs in society, ensuring that these decisions are consensual and conflicts are addressed through peaceful and democratic means.
What is interesting is that from what I can see, to value continual improving your circumstances as an individual, and setting up "fair" rules (rules for allocating costs and benefits) that a majority of the population can agree on, and ensuring safety for the population, seems to me like values that can be derived from primary rewards, and some of the very things that happiness makes us strive for.
I would love to hear what you think about this, and I look forward to reading your article on the topic.
There are variants of utilitarianism that try to account for other values besides happiness. Preference utilitarianism is probably the most famous example. Derek Parfit's "Success Theory" is the one that seems most sensible and complete to me, although I am sure there are things that it is missing.
I think the value of utilitarianism comes not from its definition of happiness as the only good, which is wrong headed. Rather, it is its view that what is good in life is something people should increase, and that actions that increase it more or for more people are proportionately more good. I think people can endorse this general ethical idea without endorsing utilitarianism.
Whatever you define as good, in normal situations all the variations of utilitarianism agree. Making sure more people are healthy, safe, economically secure, etc, maximizes pretty much any plausible formulation of the good.
You’re right that utilitarianism’s strength lies in its emphasis on increasing what is good in life for more people. This general ethical intuition resonates with many frameworks, not just utilitarianism. However, preference utilitarianism, like classical utilitarianism, faces significant challenges in practice. Without a universal metric for welfare, it becomes difficult to handle trade-offs between subjective experiences or to compare and aggregate across individuals. While these issues are abstract, they have real implications for policy-making.
I agree that in normal situations, different formulations of utilitarianism (and indeed many ethical theories) converge on similar goals—health, safety, and economic security. However, the edge cases where these formulations diverge highlight important limitations that are worth exploring. I’ll touch on contributions from theorists like Harsanyi and Binmore in later posts, as they do provide some interesting answers to these questions. I'll likely discuss Parfit’s “Success Theory” then.
If our brain just uses happiness for whatever it optimizes, than why do we use the brain to optimize for happiness? How do you reconcile this seeming duality?
If they are efficiently designed, happy feelings should reflect differences in the underlying fitness values of the options we face. Our brain seems to chase happiness because it is a guide to good decisions when picking an option.
https://www.optimallyirrational.com/p/unpacking-the-modern-science-of-happiness
https://www.optimallyirrational.com/p/unpacking-the-modern-science-of-happiness
What if we replace the word 'happiness' with the word 'contentment'? Then rather than having to continually exceed expectations one would only have to recognize it could be worse.
That would correspond to setting a low reference point. We have some agency to do so, but our hedonic system is designed to push us not to. https://www.optimallyirrational.com/p/if-you-can-you-must-why-we-set-ever
Amazing post. To your last point, Utilitarianism fails not just because evolution doesn’t care about happiness, but also because too much of it is anti-evolutionary. Dissatisfaction is an essential tool of progress. A satisfied society will always be outcompeted by a frustrated one since the former will have no reason to make scientific advances.
Chasing happiness itself is a bit like chasing grades -at all costs- in school, instead of knowledge. At some point, you start hacking the system and get good grades without learning. This would be equivalent to Huxley’s “chemically induced euphoria”. If everyone was in that state, society would crumble.
I coincidentally raised many of the same points in my latest post about Materialism. I think that focusing too much on KPIs that we can measure and improve always leads to integrating this progress as a new baseline which will neutralize the improvement. To me this happens as well on a broad societal level and not just on the personal, psychological one (that is, the hedonic treadmill).
“At some point, you start hacking the system” You are absolutely right. Any incentive system faces this challenge. Our hedonic system is designed to try to prevent this to happen. We adapt to whatever we have and become frustrated or bored, which nudges us to avoid complacency and continue aiming higher.
Excellent and thought provoking essay as always. I wondered why you didn’t consider religion though? I mean it’s obviously a collective coping mechanism for uncertainty? Religion strips people of agency & free will. We look at this now & see it as superstitious nonsense but of course it’s an incredibly powerful way to reduce anxiety / promote happiness. When you are repeatedly taught heuristics: God wills it, life is suffering, be grateful for what you have etc it is much easier to habituate than when the heuristic is: you control your own future, you are responsible for the path your life takes, anyone can win with enough effort. It must have been easier to accept relative discrepancies in wealth when you felt you had no control. But capitalism & science says: you *will* it. You *control* it. So much of modern mental health is really about reinventing the religious premise of: deal with it, accept it, it’s not your fault, look at the beautiful trees & feel grateful. It is why Buddhist teaching or self help versions of it are so popular in the West. They say, stop planning, stop wanting, stop attachments, stay in the present moment where you’re safe. Anyway, to your point about gov policy, I couldn’t agree more. But there is surely a lot to be said for crafting better collective coping strategies than state ordained medical mental health interventions. Far from reducing anxieties, these have apparently reified it into one of the big oppressors of our time. While it may be true that targetting ‘happiness’ per se is pointless, it is still the case that the loss of religion & collective / community based coping has undermined Western populations’ ability to deal with uncertainty - something that govs may wish to address?
Hi Claire, thanks for the feedback. The idea that religion is a "collective/community-based coping" mechanism is very intuitive, particularly in countries with a tradition of a "moralising god." However, the wide diversity of religious beliefs across societies and cultures suggests that religion is not necessarily such a mechanism. A great book on the topic is Boyer's "Religion Explained".
Otherwise, I think there are good alternatives to the utilitarian approach, particularly those following the contractualist approach. I plan to write a series of posts on this in the near future.
Thanks, as always, for your generous response.
Completely agree with your critical view on utilitarianism as a basis for policy. However, I don't think that Rawls' alternative works. Some sort of contractarian approach might do the job, but not his. Have you read Gauthier's Morals by Agreement? There is also Narveson's The Libertarian Idea that takes a similar approach.
Hi Max, yes I think Gauthier's perspective is more relevant to understanding how we resolve social choices. Binmore's "Natural Justice" is similarly informed by game theory and bargaining theory with an added evolutionary perspective. I will discuss it in later posts.
Of possible interest (and with apologies for the mischievous picture of utilitarians arguing). cc @Bentham's Bulldog @Richard Y Chappell @Richard Pettigrew @Lance S. Bush
“Unfortunately, utilitarianism is bound to fail. It is founded on misguided intuitions about what happiness is...This misunderstanding of happiness is a major flaw in the foundations of this philosophical approach, one that makes it unsuited to be what guides public policy.”
I have two objections to this.
First, I have an objection about the nature of happiness as understood by utilitarians. The criticism above rests on a misunderstanding of how sophisticated utilitarians conceptualise happiness. Utilitarians are not committed to a naive conception of happiness. Indeed, this is one of the reasons that contemporary utilitarians talk more in terms of overall well-being or welfare, rather than happiness. The criticism that utilitarians are seeking the repetition of an inherently transient mental state therefore misses the mark as a criticism.
Second, I have an objection about the timeframe over which utilitarians seek to maximise well-being. Utilitarianism does not commit you to the maximisation of well-being by doing, now, whatever would increase well-being most, now. This has long been understood as a flawed way of maximising well-being over the timescale of a human life. Utilitarianism is consistent with forms of suffering that are conducive to increases in overall well-being, like training to become a soldier or undergoing a PhD. Longtermist utilitarians even go well beyond the timescale of a single human life and seek to maximise well-being over the very, very long term.
Hi Michael, thanks for your comment. Here are answers:
“First, I have an objection about the nature of happiness as understood by utilitarians. The criticism above rests on a misunderstanding of how sophisticated utilitarians conceptualise happiness. Utilitarians are not committed to a naive conception of happiness. Indeed, this is one of the reasons that contemporary utilitarians talk more in terms of overall well-being or welfare, rather than happiness. The criticism that utilitarians are seeking the repetition of an inherently transient mental state therefore misses the mark as a criticism.”
- My criticism is mostly targeted at classical utilitarianism, which is the original and most intuitive version of the approach. Modern utilitarians, like Peter Singer, often adopt a “preference” framework, partly to address the conceptual problems ‘of the classical approach. I pre-empted this point in a footnote in the post. This conceptual shift introduces deeper issues. Once happiness is no longer treated as a measurable and aggregable quantity, the objective function of utilitarianism becomes unclear. What exactly is being maximised, and how should trade-offs be handled? This creates a more complex, arguably less coherent framework.
“Second, I have an objection about the timeframe over which utilitarians seek to maximise well-being. Utilitarianism does not commit you to the maximisation of well-being by doing, now, whatever would increase well-being most, now. This has long been understood as a flawed way of maximising well-being over the timescale of a human life. Utilitarianism is consistent with forms of suffering that are conducive to increases in overall well-being, like training to become a soldier or undergoing a PhD. Longtermist utilitarians even go well beyond the timescale of a single human life and seek to maximise well-being over the very, very long term.”
- That is a fair point, but this approach requires principles for adjudicating trade-offs between short-term and long-term well-being. If classical utilitarianism is set aside, it becomes unclear what principles should replace the maximisation of happiness as an experienced hedonic feeling.
Thank you very much for your reply, Lionel.
I will add some further thoughts, but will keep them relatively short to avoid a prolonged exchange (on the assumption this is not desirable).
Sticking with the language of well-being, rather than happiness, it is quite clear that this can be measured and aggregated to some extent. Well-being comes in degrees, some people have higher levels than others, and at least some of these degrees of difference can be readily recognised. It is still well-being that is being maximised overall. The phenomenon of habituation and relative comparison makes our understanding of well-being more sophisticated and tells us that we cannot simply add more units of, say, wealth for more corresponding units of well-being. Problem cases and precise formulations are a problem for academics, but policy makers could settle on some generally agreed upon metrics with broad consensus.
Perhaps a more interesting and controversial question is how commensurable differing degrees of well-being are and how these can be precisely formulated (for example, should they be understood as one unit along one spectrum, or multiple units along multiple spectrums? Is this the same for all species, or is it species-relative?)
Where you say the framework becomes less coherent, I am inclined to say that it simply becomes more sophisticated. However, I do suspect that the more sophisticated the framework becomes the less clearly action-guiding it becomes. This is a problem for a moral theory if it doesn’t just want to offer a criterion of rightness, but also wants to offer moral agents a decision-making procedure. This touches directly on your second comment, which I agree is a key area for discussion. Even if we accept the maximisation of well-being as the criterion of rightness, how does the sophisticated utilitarian make decisions to adjudicate trade-offs?
You say, "it is quite clear that this can be measured and aggregated to some extent." I think this is actually a very complex question. Whether it is measurable within an individual has been at the centre of debates in economics and psychology for decades. See, for instance, this book for a historical perspective:
https://www.amazon.com/Measuring-Utility-Revolution-Behavioral-Economics/dp/0199372772
My take is that modern neuroscience and evolutionary reasoning suggest that happiness, well-being, or utility does not behave like a quantity that can be measured.
Assuming it could be measured, there would still be the problem of aggregation: how to add units of well-being across different individuals. Even for utilitarians, this is a major issue.
Utilitarians who opt for a "preference" approach are, in my view, akin to economists adopting an ordinalist position about utility (utility is just representing preferences for some things over other things). This makes aggregation conceptually impossible because there is no underlying quantity to measure and aggregate behind the preferences.
For a quick glimpse at my more fundamental take on utilitarianism:
I think utilitarianism is intuitive at some level because we can use it in certain social situations. Binmore, in his book "Natural Justice", points out that utilitarianism is feasible in social contracts where there is external enforcement of the rules of the social contract (e.g., children can accept utilitarianism enforced by parental rules). However, at the societal level, there is no external enforcement—there is nothing above society to enforce a social contract. Therefore, a utilitarian contract is always bound to be contested and broken by dissatisfied parties. It cannot be a stable solution. I will write about this at length in future posts.
Thank you for an engaging and generous exchange, Lionel. You’ve given me lots of useful points to consider.
I’m still of the view that overall well-being comes in degrees, that this varies within and between individuals, and that such variation can be understood in terms of other (perhaps first order) qualities, such as being in better or worse health. But I’m also more sympathetic to some of the complexities you have raised.
Many thanks for the book recommendation, too. Most of my current books on ‘utility’ are works of philosophy, but I’d be very interested to add works from economics and psychology. An overview, such as the one you mention, fits the bill perfectly.
I will subscribe to your page and I forward to reading more of your posts in future. Wishing you all the best.
This is a problem with how happiness is measured (with reference to an ordinal scale), not a problem with happiness. As I've said before it's like measuring height by asking children whether they are tall, medium or short. You'll get much the same distribution whether you ask them at age 5 or at age 15. So, you could conclude that even though some children are taller than others, no one gets any taller over time.
The hedonic treadmill is as you well describe, but your characterization of utilitarianism is skewed similarly to how opponents’ characterizations of deontology include rigid and absurd adherence to rules even when a particular situation clearly calls for deviation. Many utilitarians, counter to how you represent them, seem at times overly concerned with future people/consequences (climate change, AI misalignment, etc.) at the expense of current people and happiness.
I’m not a utilitarian or a deontologist or a pragmatist or anything else because I see the obvious strengths all these philosophies offer depending on circumstance. All are insightful, and the danger lies in dogmatically accepting or rejecting any one of them at the expense of all others. This of course leads to extremism, which is the one philosophy I see no strength in.
This is a very timely warning for Indian policy makers. The recent spurt in Direct Benefit Transfers right before State Assembly elections will likely be transformed into a spiral of ever-increasing amounts for the beneficiaries-thanks to habituation.
We need to think of potent measures to safeguard our country's fiscal health.
A measurable quantity has specific properties. Length and volume are examples of measurable quantities. Classical utilitarianism treat happiness as such a thing: a quantity that can be accumulated. Happiness is not something like that. Due to habituation, increases in happiness will tend to fade away.
The answer is in the post. Improving conditions at the bottom end has an effect. Not so much for higher levels. What to do is for another post. But there are alternatives to utilitarianism. The most influential one is Rawls' approach. It is likely supported by more political philosophers than utilitarianism.
The non linearity, yes to some extent, though it is more extreme than typically assumed in the utilitarian literature. In addition, habituation and the illusion that habituation will not happen are major challenges to utilitarianism (recognised as such for quite some time).