His point was about interpersonal comparison, which is a different issue from cardinal vs ordinal. Utility could be ordinal but interpersonally comparable or cardinal but not.
Here's a plausible principle: the difference in harm between torturing me to death and poking me with a sharp stick is greater than the difference in harm between lightly prodding me and poking me with a sharp stick. If utility is merely ordinal, then we can't make this judgment. Yet this judgment is obviously true. There are, of course, other reasons to adopt cardinal utility, some of which David discusses in the article.
Of course, some people (masochists) enjoy torture and deliberately seek it out. While almost certainly they do not seek “torture to death”, it’s plausible in a heightened mental state or trance they accept death or even welcome it. I understand the temptation to seek cardinal utility but it just looks like wishful thinking and a desire to control the behavior of others “for their own benefit”. Subjectivity is an insurmountable wall here.
They seek it out because it brings them pleasure and it doesn't cause them to experience actually undesirable mental states (suffering). Maybe they experience pain, but it isn't suffering if it doesn't bother them and they like it.
I generally am against controlling the behavior of others for their own benefit--I think that history shows that tends to be ineffective. But we mustn't allow utility functions to be politicized--the mere fact that there are better reasons to be a libertarian if interpersonal utility isn't comparable and utility is ordinal doesn't say anything about what is true.
I don't think it's wishful thinking--as I said before, it's the only way to allow one to make decisions under uncertainty with a coherent utility function, as Von Neuman and Morgenstern showed, and it's the only way to make comparative judgments like "the difference in harm between me eating a tomato when I don't like tomatoes much and me being viciously tortured to death is greater than the difference in harm between another person eating a food they don't like and eating a different food that they dislike slightly more."
I believe the original presentation was in an appendix to the second edition of _The Theory of Games and Economic Behavior_. There may be later and better explanations, by them or by other people.
Strictly speaking, they didn't. They showed that if their 4 axioms held for agents, a utility function existed. But maybe I am being too picky about language. Presumably, if we deny any of the axioms, then the utility function might not exist. One obviously can make decisions a different way. Perhaps DF implicitly means that there is not other way to make decisions that are rational. But even that seems a bit strong, as one might discover a different method that duplicated the results exactly.
I’m not sure how you get from the cardinality implied by VNM to interpersonal comparisons. VNM shows that the function exists, but it is “ uniquely determined (up to adding a constant and multiplying by a positive scalar)”. So assuming we somehow discovered various persons' utility functions, I am not sure how we would calibrate them onto the same scale. We still need a unit of measure that is common to all, so we know what constant to add or scalar to multiply by. Maybe the difference between unending torture and orgasm would be comparable in all persons? Wikipedia says, “ VNM-utility is not canonically additive across multiple individuals (see Limitations), so "total VNM-utility" and "average VNM-utility" are not immediately meaningful (some sort of normalization assumption is required).”
The Wikipedia article I am cribbing from also points out the difference between decision utility (basis for evaluating and making decisions ex ante) and experience utility (how much you enjoy or hate what is happening as it happens). Ideally utilitarian maximization would synchronize them, so that decision utility maximization would lead to experience utility maximization, but I think that needs a raft of additional assumptions about our understanding of cause and effect in addition to individuals' utility functions (of either or both kinds).
What precisely is being compared there, the utility change of the effects, or the loss of life that usually follows from setting someone on fire, compared to the usually non-fatal stick poking? If the poke is also fatal, how would one compare them? Is the utility loss worse if the burn victim dies or if they survive but experience months of torture trying to recover, and never recover fully?
What is the standard unit we can use to measure the difference? What instrument can we use to measure it? How many dimensions does it have?
If we stipulate that large differences are comparable, does it follow that small differences must also be comparable? Because I doubt that this sort of large comparison is what people usually want to make when they treat utility as cardinal.
Perhaps we should not say interpersonal utility is not comparable, but that there are too many standards by which to compare it, and no consensus on which one should take priority. Then we might say that utility is comparable, without being measurable. But then it begins to sound like an ordinal concept instead of a cardinal one.
By the way, thanks for recommending this Substack on ACX, I followed your clue to get here.
//What precisely is being compared there, the utility change of the effects, or the loss of life that usually follows from setting someone on fire//
The utility change. We can stipulate that no one dies.
// Is the utility loss worse if the burn victim dies or if they survive but experience months of torture trying to recover, and never recover fully?//
This would be an empirical question.
//What is the standard unit we can use to measure the difference? //
Happiness.
//What instrument can we use to measure it?//
We don't have great instruments, but we can use our reflective judgments, just as we can guess that I'm experiencing brighter quales than someone who is in a perfectly black room. Though I'd imagine full information about neuroscience, combined with knowledge of the psychophysical laws could allow us to know what things actually maximize utility.
//How many dimensions does it have?//
I don't know what this means.
//If we stipulate that large differences are comparable, does it follow that small differences must also be comparable? Because I doubt that this sort of large comparison is what people usually want to make when they treat utility as cardinal.//
Yes it would. However, the argument for interpersonal comparisons of utility is, among other things, that if it were impossible, I couldn't say that someone was harmed more than someone else. But this is clearly absurd. Note, here I was only expressing my belief in utility being interpersonally comparable, though I do also think it's cardinal.
//Perhaps we should not say interpersonal utility is not comparable, but that there are too many standards by which to compare it, and no consensus on which one should take priority.//
There are different things that might make a person well off. But once we specify what makes a person well off--in this case, generally happiness--it will be interpersonally measurable.
//Then we might say that utility is comparable, without being measurable. But then it begins to sound like an ordinal concept instead of a cardinal one.//
Here's a plausible principle: the difference in harm between torturing me to death and poking me with a sharp stick is greater than the difference in harm between lightly prodding me and poking me with a sharp stick. If utility is merely ordinal, then we can't make this judgment. Yet this judgment is obviously true.
//Happiness [is the standard unit we can use to measure the difference] //
How do we define a standard unit of happiness?
//We can guess that I'm experiencing brighter quales than someone who is in a perfectly black room.//
Should we conclude that quales are also interpersonally comparable or cardinal? That they are unidimensional? Analogies are good for clarifying an idea, but questionable as actual arguments.
//I don't know what [ How many dimensions does it have? ] means. //
It means that pain, harm, death, sympathy, and other negative physical, social, emotional, or economic experiences can vary independently, but are considered to affect happiness to some degree. Perhaps if each of those were cardinal or interpersonally comparable, we could combine them into a composite score of happiness, reducing the many dimensions to one. But this seems to involve a sort of judgement about which different persons might legitimately disagree.
And if so, happiness misleads as a label. Some stoic persons who have experienced great hardship remain happy, and some healthy trust fund kids are miserable.
This relates to a criticism by Philipa Foot, that utilitarianism tries to reduce ethics to the virtue of benevolence. But other virtues are not reducible to benevolence. This means sometimes they can come into conflict, creating moral dilemmas. To reduce goodness to a single dimension assumes that there is a constant tradeoff between one kind and another.
To marginalists, this should seem odd. In economics, exchange allows us to “intersubjectivize” many subjective evaluations into a market price, which is not constant from time to time or place to place. How do we intersubjectivize non-market experiences, and why should we expect it to be more stable than prices?
Perhaps everyone is very similar, and the basic parameters of the decision are fairly constant. Hmmm…
//However, the argument for interpersonal comparisons of utility is, among other things, that if it were impossible, I couldn't say that someone was harmed more than someone else. //
Harm is still difficult to quantify, but note that it is a more narrowly defined dimension than happiness or utility. Other variables can affect utility independently. This changed the subject from the multidimensional concept to one of its perhaps unidimensional components. If harm is comparable, it doesn’t follow that utility or happiness is, unless we treat them as synonyms and treat harm itself as multi-dimensional. But then we still lack the technique for reducing those dimensions into one cardinal dimension.
//There are different things that might make a person well off. But once we specify what makes a person well off--in this case, generally happiness--it will be interpersonally measurable. //
“Well off” is another related term. Should we treat it as perfectly synonymous to utility? Or a component dimension?
//the difference in harm between torturing me to death and poking me with a sharp stick is greater than the difference in harm between lightly prodding me and poking me with a sharp stick. If utility is merely ordinal, then we can't make this judgment. Yet this judgment is obviously true. //
Again, this compares harms instead of utilities. This holds other variables constant while varying a (presumably) single variable. But that variable is not utility, it is a component of utility, an input into the utility function. It is possible for all the physical experiences that affect utility to be physically measurable without saying anything necessary about how the subject evaluates or experiences their combined effects.
//How do we define a standard unit of happiness?//
The units are irrelevant. Mass is objective, but you can have the units for mass be whatever you want.
//Should we conclude that quales are also interpersonally comparable or cardinal? That they are unidimensional? Analogies are good for clarifying an idea, but questionable as actual arguments.//
They'd be cardinal.
//And if so, happiness misleads as a label. Some stoic persons who have experienced great hardship remain happy, and some healthy trust fund kids are miserable.
This relates to a criticism by Philipa Foot, that utilitarianism tries to reduce ethics to the virtue of benevolence. But other virtues are not reducible to benevolence. This means sometimes they can come into conflict, creating moral dilemmas. To reduce goodness to a single dimension assumes that there is a constant tradeoff between one kind and another.//
There are different things that make people happy, but there is a fact about how happy a person is.
//Harm is still difficult to quantify, but note that it is a more narrowly defined dimension than happiness or utility. Other variables can affect utility independently. This changed the subject from the multidimensional concept to one of its perhaps unidimensional components.//
But happiness is a brain state--it's measurable in principle. Why is it possible to compare harm but not utility?
//“Well off” is another related term. Should we treat it as perfectly synonymous to utility? Or a component dimension?//
There are different views about what makes a person well off. I think most of them will be comparable.
//The units are irrelevant. Mass is objective, but you can have the units for mass be whatever you want. //
So for mass, you can pick an arbitrary object and use it as your standard unit. You can use a scale to compare reasonable scale objects to it. How would we do something analogous for the util?
//[ quales would] be cardinal. //
And there is no problem there? Does the distinction between quantity and quality describe anything?
//There are different things that make people happy, but there is a fact about how happy a person is. //
But utilitarians have no access to how happy a person is except through the things that make them happy (Or don’t). Maybe I have misunderstood. Is utilitarianism about enabling people to adjust their own level of happiness, or about making decisions according to a formula that is supposed to take their happiness into account, but doesn’t actually consult them, because utility is cardinal?
//But happiness is a brain state--it's measurable in principle. Why is it possible to compare harm but not utility? //
Perhaps it is not, since harm may also be inherently multi-dimensional. If we assume harm is uni-dimensional, perhaps we could find some acceptable unit of measure. Nothing suggests itself to me, but my incredulity is not an argument. Neither is emphatic assertion.
////“Well off” is another related term. Should we treat it as perfectly synonymous to utility? Or a component dimension?////
//There are different views about what makes a person well off. I think most of them will be comparable.//
Most of them?
Yes, “well off” is easy. How much money do you have? But that is not utility. Or we could treat “well off” as a synonym of utility, in which case it’s use is rhetorical. It is a figure of speech that is used ordinarily in a very loose way, and has connotations that utility does not, or at least should not have.
How do we reduce a multi-dimensional space to a single dimension without losing information? It is only possible if the various initial dimensions are basically redundant. In practice, we are compelled to do this all the time, but we can use all our background knowledge to evaluate the multiple dimensions and reduce it to (do/don't do.).
Actual utility is an agent-relative experience. It’s not comparable across parties since it’s entirely subjective. Although there are certain extremes that wouldn’t fall into anyone’s utility function, society can’t make public policy based on subjective states. Pursuit of utility is within the domain of the individual and their close personal relationships.
This argument instead makes the case for maximizing certain objective factors, or primary goods, that tend to put people in a better position to maximize their own subjective well-being. These should be increased regardless of one’s subjective preference for them.
No one would agree to maximizing someone else’s happiness because who knows what makes them happy. But people would agree to maximizing a collective pot of generally useful resources.
I like your point, though I may be projecting my thoughts onto it.
Usually utilitarian arguments proceed as if a benevolent dictator was considering options. This policy promises to bring the most utility, so it should be chosen. What is the alternative, do you want the second best or what?
A different perspective would allow for different persons trying different things, and their projects succeeding on the basis of whether the participants think it improves their lives. Then the utils measure themselves, and don’t need to be compared interpersonally.
Maybe that overlooks some important point? Where is interpersonally comparable utility needed? Maybe it only matters when we are going to exclude alternatives. We could save lots of resources if we skip to the end of the discovery process, but are we that smart?
Or maybe we want (paternalistically) to prevent persons with different evaluations from making terrible mistakes, but we also don’t want to empower demagogues to put their thumb on the utility scale. Does interpersonal comparison of utility help us with that? I guess it would. We could just slap our utilometer on someone and show that the demagogue was full of crap.
Sounds nice, but when do we expect a reliable utilometer to come on the market? Does it only measure present utility, or future utility?
Thanks for the response. I discuss the idea of a benevolent dictator maximizing utility below. Summary: say you can have someone forced into maximizing their utility. If you wouldn’t force this on another person (despite having a duty to do so under utilitarianism), then you are valuing that persons freedom independently and above the value placed on their welfare. This is (a reason) that I argue that freedom is the foundation of morality.
On interpersonal utility comparisons, consider gift giving. Gifts other than, for example, cash transfer payments are often derided by economists are producing deadweight loss. A great gift, however, is one which signals that the gift-giver has a refined sense of the gift-getter's preferences. Particularly given uncertainty and agency costs, it easy to tell a story in which this signal would be more valuable in absolute terms than the deadweight loss of the gift.
The best point is the risk idea. But I am inclined to push back.
My initial, perhaps quite biased and motivated-reasoning response, is that it is classic “the world has to be this way, because I can’t think of a model that doesn’t work like that.” It is very handy to analyze circumstances in a particular way, but this handiness doesn’t mean that this is what everyone consciously does. If we do it unconsciously, that is an interesting fact about evolution. If we don’t actually do it consistently, consciously or otherwise, what does the model tell us? Is it positive or normative?
If it is positive, supposedly we can test it and possibly discover it is wrong. (The behavioral economists think they have done so, though maybe they are mistaken.)
If it is normative, we are back where we started. Having a model of how people ought to respond to risk would say nothing about whether utility is interpersonally comparable; it would say only that we have to treat circumstances that way to use this neat model.
The risk model is very appropriate when the subject knows all the variables and they are all cardinal values that can be estimated or measured. The fact that different circumstances would invalidate the model does not mean that circumstances cannot be like that.
The risk model assumes we are at an extreme end of the explore/exploit continuum, where the only unknowns are the actual values of random variables. And utility ideally would be one of those known or estimable quantities. But that is just looking for the keys where the light is best. How do we know that is what we are looking at, rather than what we wish to see? It doesn’t convince me that utility is measurable as a cardinal quantity or interpersonally comparable in a determinate way.
Assuming you are a benevolent dictator, or even a economic advisor with a reasonable degree of certainty that your advice will be implemented. Would you try to compensate the losers of a economic improvement, say through transfers in wealth or a reallocation of rights. Assuming the deadweight loss from taxation and such was minimal. From what I can tell this seems to be something Robin Hanson would advise.
There is a problem getting people to reveal their true utility function if people get paid for lying. I vaguely recall a scheme some economist came up with to get people to reveal their true preferences, but I have forgotten the details.
It's always been obvious to me that utility is interpersonally comparable. Setting a person on fire is worse than poking someone else with a stick is.
This is ordinal utility, not cardinal. How much worse is it? There is no way to assign a value.
His point was about interpersonal comparison, which is a different issue from cardinal vs ordinal. Utility could be ordinal but interpersonally comparable or cardinal but not.
Here's a plausible principle: the difference in harm between torturing me to death and poking me with a sharp stick is greater than the difference in harm between lightly prodding me and poking me with a sharp stick. If utility is merely ordinal, then we can't make this judgment. Yet this judgment is obviously true. There are, of course, other reasons to adopt cardinal utility, some of which David discusses in the article.
Of course, some people (masochists) enjoy torture and deliberately seek it out. While almost certainly they do not seek “torture to death”, it’s plausible in a heightened mental state or trance they accept death or even welcome it. I understand the temptation to seek cardinal utility but it just looks like wishful thinking and a desire to control the behavior of others “for their own benefit”. Subjectivity is an insurmountable wall here.
They seek it out because it brings them pleasure and it doesn't cause them to experience actually undesirable mental states (suffering). Maybe they experience pain, but it isn't suffering if it doesn't bother them and they like it.
I generally am against controlling the behavior of others for their own benefit--I think that history shows that tends to be ineffective. But we mustn't allow utility functions to be politicized--the mere fact that there are better reasons to be a libertarian if interpersonal utility isn't comparable and utility is ordinal doesn't say anything about what is true.
I don't think it's wishful thinking--as I said before, it's the only way to allow one to make decisions under uncertainty with a coherent utility function, as Von Neuman and Morgenstern showed, and it's the only way to make comparative judgments like "the difference in harm between me eating a tomato when I don't like tomatoes much and me being viciously tortured to death is greater than the difference in harm between another person eating a food they don't like and eating a different food that they dislike slightly more."
I believe the original presentation was in an appendix to the second edition of _The Theory of Games and Economic Behavior_. There may be later and better explanations, by them or by other people.
Strictly speaking, they didn't. They showed that if their 4 axioms held for agents, a utility function existed. But maybe I am being too picky about language. Presumably, if we deny any of the axioms, then the utility function might not exist. One obviously can make decisions a different way. Perhaps DF implicitly means that there is not other way to make decisions that are rational. But even that seems a bit strong, as one might discover a different method that duplicated the results exactly.
I’m not sure how you get from the cardinality implied by VNM to interpersonal comparisons. VNM shows that the function exists, but it is “ uniquely determined (up to adding a constant and multiplying by a positive scalar)”. So assuming we somehow discovered various persons' utility functions, I am not sure how we would calibrate them onto the same scale. We still need a unit of measure that is common to all, so we know what constant to add or scalar to multiply by. Maybe the difference between unending torture and orgasm would be comparable in all persons? Wikipedia says, “ VNM-utility is not canonically additive across multiple individuals (see Limitations), so "total VNM-utility" and "average VNM-utility" are not immediately meaningful (some sort of normalization assumption is required).”
The Wikipedia article I am cribbing from also points out the difference between decision utility (basis for evaluating and making decisions ex ante) and experience utility (how much you enjoy or hate what is happening as it happens). Ideally utilitarian maximization would synchronize them, so that decision utility maximization would lead to experience utility maximization, but I think that needs a raft of additional assumptions about our understanding of cause and effect in addition to individuals' utility functions (of either or both kinds).
https://en.m.wikipedia.org/wiki/Von_Neumann%E2%80%93Morgenstern_utility_theorem
What precisely is being compared there, the utility change of the effects, or the loss of life that usually follows from setting someone on fire, compared to the usually non-fatal stick poking? If the poke is also fatal, how would one compare them? Is the utility loss worse if the burn victim dies or if they survive but experience months of torture trying to recover, and never recover fully?
What is the standard unit we can use to measure the difference? What instrument can we use to measure it? How many dimensions does it have?
If we stipulate that large differences are comparable, does it follow that small differences must also be comparable? Because I doubt that this sort of large comparison is what people usually want to make when they treat utility as cardinal.
Perhaps we should not say interpersonal utility is not comparable, but that there are too many standards by which to compare it, and no consensus on which one should take priority. Then we might say that utility is comparable, without being measurable. But then it begins to sound like an ordinal concept instead of a cardinal one.
By the way, thanks for recommending this Substack on ACX, I followed your clue to get here.
//What precisely is being compared there, the utility change of the effects, or the loss of life that usually follows from setting someone on fire//
The utility change. We can stipulate that no one dies.
// Is the utility loss worse if the burn victim dies or if they survive but experience months of torture trying to recover, and never recover fully?//
This would be an empirical question.
//What is the standard unit we can use to measure the difference? //
Happiness.
//What instrument can we use to measure it?//
We don't have great instruments, but we can use our reflective judgments, just as we can guess that I'm experiencing brighter quales than someone who is in a perfectly black room. Though I'd imagine full information about neuroscience, combined with knowledge of the psychophysical laws could allow us to know what things actually maximize utility.
//How many dimensions does it have?//
I don't know what this means.
//If we stipulate that large differences are comparable, does it follow that small differences must also be comparable? Because I doubt that this sort of large comparison is what people usually want to make when they treat utility as cardinal.//
Yes it would. However, the argument for interpersonal comparisons of utility is, among other things, that if it were impossible, I couldn't say that someone was harmed more than someone else. But this is clearly absurd. Note, here I was only expressing my belief in utility being interpersonally comparable, though I do also think it's cardinal.
//Perhaps we should not say interpersonal utility is not comparable, but that there are too many standards by which to compare it, and no consensus on which one should take priority.//
There are different things that might make a person well off. But once we specify what makes a person well off--in this case, generally happiness--it will be interpersonally measurable.
//Then we might say that utility is comparable, without being measurable. But then it begins to sound like an ordinal concept instead of a cardinal one.//
Here's a plausible principle: the difference in harm between torturing me to death and poking me with a sharp stick is greater than the difference in harm between lightly prodding me and poking me with a sharp stick. If utility is merely ordinal, then we can't make this judgment. Yet this judgment is obviously true.
You're welcome for recommending the blog!
//Happiness [is the standard unit we can use to measure the difference] //
How do we define a standard unit of happiness?
//We can guess that I'm experiencing brighter quales than someone who is in a perfectly black room.//
Should we conclude that quales are also interpersonally comparable or cardinal? That they are unidimensional? Analogies are good for clarifying an idea, but questionable as actual arguments.
//I don't know what [ How many dimensions does it have? ] means. //
It means that pain, harm, death, sympathy, and other negative physical, social, emotional, or economic experiences can vary independently, but are considered to affect happiness to some degree. Perhaps if each of those were cardinal or interpersonally comparable, we could combine them into a composite score of happiness, reducing the many dimensions to one. But this seems to involve a sort of judgement about which different persons might legitimately disagree.
And if so, happiness misleads as a label. Some stoic persons who have experienced great hardship remain happy, and some healthy trust fund kids are miserable.
This relates to a criticism by Philipa Foot, that utilitarianism tries to reduce ethics to the virtue of benevolence. But other virtues are not reducible to benevolence. This means sometimes they can come into conflict, creating moral dilemmas. To reduce goodness to a single dimension assumes that there is a constant tradeoff between one kind and another.
To marginalists, this should seem odd. In economics, exchange allows us to “intersubjectivize” many subjective evaluations into a market price, which is not constant from time to time or place to place. How do we intersubjectivize non-market experiences, and why should we expect it to be more stable than prices?
Perhaps everyone is very similar, and the basic parameters of the decision are fairly constant. Hmmm…
//However, the argument for interpersonal comparisons of utility is, among other things, that if it were impossible, I couldn't say that someone was harmed more than someone else. //
Harm is still difficult to quantify, but note that it is a more narrowly defined dimension than happiness or utility. Other variables can affect utility independently. This changed the subject from the multidimensional concept to one of its perhaps unidimensional components. If harm is comparable, it doesn’t follow that utility or happiness is, unless we treat them as synonyms and treat harm itself as multi-dimensional. But then we still lack the technique for reducing those dimensions into one cardinal dimension.
//There are different things that might make a person well off. But once we specify what makes a person well off--in this case, generally happiness--it will be interpersonally measurable. //
“Well off” is another related term. Should we treat it as perfectly synonymous to utility? Or a component dimension?
//the difference in harm between torturing me to death and poking me with a sharp stick is greater than the difference in harm between lightly prodding me and poking me with a sharp stick. If utility is merely ordinal, then we can't make this judgment. Yet this judgment is obviously true. //
Again, this compares harms instead of utilities. This holds other variables constant while varying a (presumably) single variable. But that variable is not utility, it is a component of utility, an input into the utility function. It is possible for all the physical experiences that affect utility to be physically measurable without saying anything necessary about how the subject evaluates or experiences their combined effects.
//How do we define a standard unit of happiness?//
The units are irrelevant. Mass is objective, but you can have the units for mass be whatever you want.
//Should we conclude that quales are also interpersonally comparable or cardinal? That they are unidimensional? Analogies are good for clarifying an idea, but questionable as actual arguments.//
They'd be cardinal.
//And if so, happiness misleads as a label. Some stoic persons who have experienced great hardship remain happy, and some healthy trust fund kids are miserable.
This relates to a criticism by Philipa Foot, that utilitarianism tries to reduce ethics to the virtue of benevolence. But other virtues are not reducible to benevolence. This means sometimes they can come into conflict, creating moral dilemmas. To reduce goodness to a single dimension assumes that there is a constant tradeoff between one kind and another.//
There are different things that make people happy, but there is a fact about how happy a person is.
//Harm is still difficult to quantify, but note that it is a more narrowly defined dimension than happiness or utility. Other variables can affect utility independently. This changed the subject from the multidimensional concept to one of its perhaps unidimensional components.//
But happiness is a brain state--it's measurable in principle. Why is it possible to compare harm but not utility?
//“Well off” is another related term. Should we treat it as perfectly synonymous to utility? Or a component dimension?//
There are different views about what makes a person well off. I think most of them will be comparable.
//The units are irrelevant. Mass is objective, but you can have the units for mass be whatever you want. //
So for mass, you can pick an arbitrary object and use it as your standard unit. You can use a scale to compare reasonable scale objects to it. How would we do something analogous for the util?
//[ quales would] be cardinal. //
And there is no problem there? Does the distinction between quantity and quality describe anything?
//There are different things that make people happy, but there is a fact about how happy a person is. //
But utilitarians have no access to how happy a person is except through the things that make them happy (Or don’t). Maybe I have misunderstood. Is utilitarianism about enabling people to adjust their own level of happiness, or about making decisions according to a formula that is supposed to take their happiness into account, but doesn’t actually consult them, because utility is cardinal?
//But happiness is a brain state--it's measurable in principle. Why is it possible to compare harm but not utility? //
Perhaps it is not, since harm may also be inherently multi-dimensional. If we assume harm is uni-dimensional, perhaps we could find some acceptable unit of measure. Nothing suggests itself to me, but my incredulity is not an argument. Neither is emphatic assertion.
////“Well off” is another related term. Should we treat it as perfectly synonymous to utility? Or a component dimension?////
//There are different views about what makes a person well off. I think most of them will be comparable.//
Most of them?
Yes, “well off” is easy. How much money do you have? But that is not utility. Or we could treat “well off” as a synonym of utility, in which case it’s use is rhetorical. It is a figure of speech that is used ordinarily in a very loose way, and has connotations that utility does not, or at least should not have.
How do we reduce a multi-dimensional space to a single dimension without losing information? It is only possible if the various initial dimensions are basically redundant. In practice, we are compelled to do this all the time, but we can use all our background knowledge to evaluate the multiple dimensions and reduce it to (do/don't do.).
Actual utility is an agent-relative experience. It’s not comparable across parties since it’s entirely subjective. Although there are certain extremes that wouldn’t fall into anyone’s utility function, society can’t make public policy based on subjective states. Pursuit of utility is within the domain of the individual and their close personal relationships.
This argument instead makes the case for maximizing certain objective factors, or primary goods, that tend to put people in a better position to maximize their own subjective well-being. These should be increased regardless of one’s subjective preference for them.
No one would agree to maximizing someone else’s happiness because who knows what makes them happy. But people would agree to maximizing a collective pot of generally useful resources.
I like your point, though I may be projecting my thoughts onto it.
Usually utilitarian arguments proceed as if a benevolent dictator was considering options. This policy promises to bring the most utility, so it should be chosen. What is the alternative, do you want the second best or what?
A different perspective would allow for different persons trying different things, and their projects succeeding on the basis of whether the participants think it improves their lives. Then the utils measure themselves, and don’t need to be compared interpersonally.
Maybe that overlooks some important point? Where is interpersonally comparable utility needed? Maybe it only matters when we are going to exclude alternatives. We could save lots of resources if we skip to the end of the discovery process, but are we that smart?
Or maybe we want (paternalistically) to prevent persons with different evaluations from making terrible mistakes, but we also don’t want to empower demagogues to put their thumb on the utility scale. Does interpersonal comparison of utility help us with that? I guess it would. We could just slap our utilometer on someone and show that the demagogue was full of crap.
Sounds nice, but when do we expect a reliable utilometer to come on the market? Does it only measure present utility, or future utility?
Thanks for the response. I discuss the idea of a benevolent dictator maximizing utility below. Summary: say you can have someone forced into maximizing their utility. If you wouldn’t force this on another person (despite having a duty to do so under utilitarianism), then you are valuing that persons freedom independently and above the value placed on their welfare. This is (a reason) that I argue that freedom is the foundation of morality.
https://open.substack.com/pub/neonomos/p/the-utility-coach-thought-experiment?r=1pded0&utm_medium=ios&utm_campaign=post
On interpersonal utility comparisons, consider gift giving. Gifts other than, for example, cash transfer payments are often derided by economists are producing deadweight loss. A great gift, however, is one which signals that the gift-giver has a refined sense of the gift-getter's preferences. Particularly given uncertainty and agency costs, it easy to tell a story in which this signal would be more valuable in absolute terms than the deadweight loss of the gift.
The best point is the risk idea. But I am inclined to push back.
My initial, perhaps quite biased and motivated-reasoning response, is that it is classic “the world has to be this way, because I can’t think of a model that doesn’t work like that.” It is very handy to analyze circumstances in a particular way, but this handiness doesn’t mean that this is what everyone consciously does. If we do it unconsciously, that is an interesting fact about evolution. If we don’t actually do it consistently, consciously or otherwise, what does the model tell us? Is it positive or normative?
If it is positive, supposedly we can test it and possibly discover it is wrong. (The behavioral economists think they have done so, though maybe they are mistaken.)
If it is normative, we are back where we started. Having a model of how people ought to respond to risk would say nothing about whether utility is interpersonally comparable; it would say only that we have to treat circumstances that way to use this neat model.
The risk model is very appropriate when the subject knows all the variables and they are all cardinal values that can be estimated or measured. The fact that different circumstances would invalidate the model does not mean that circumstances cannot be like that.
The risk model assumes we are at an extreme end of the explore/exploit continuum, where the only unknowns are the actual values of random variables. And utility ideally would be one of those known or estimable quantities. But that is just looking for the keys where the light is best. How do we know that is what we are looking at, rather than what we wish to see? It doesn’t convince me that utility is measurable as a cardinal quantity or interpersonally comparable in a determinate way.
Assuming you are a benevolent dictator, or even a economic advisor with a reasonable degree of certainty that your advice will be implemented. Would you try to compensate the losers of a economic improvement, say through transfers in wealth or a reallocation of rights. Assuming the deadweight loss from taxation and such was minimal. From what I can tell this seems to be something Robin Hanson would advise.
There is a problem getting people to reveal their true utility function if people get paid for lying. I vaguely recall a scheme some economist came up with to get people to reveal their true preferences, but I have forgotten the details.