Value, Utility, and Justification
by Stillwater
In a recent post which is now unfortunately closed (“Something, Something … Petard”), James Hanley and I (and Shazbot!) engaged in a friendly dispute ostensibly about the meaning of the word “value” in economic contexts. The focal point of the discussion was a statement of mine regarding what I took to be the only relevant distinction between two types of individual actions. The first type of action is a cash exchange for a product at an amount which is so small as to be effectively meaningless to the producing firm’s survivability. The other is casting a vote for a candidate which in turn is also effectively meaningless to determining the outcome of an election. [This isn’t really relevant to the remainder of the post and apologies if this is confusing to anyone.]
In comments, I wrote that the only relevant distinction between the two scenarios is that “there’s an objective exchange of value in one and not the other. But it seems like all the libertarians on this site seem to take great pleasure in take great pleasure in explaining to me that value isn‘t objectively measurable”. Hanley’s response was that I don’t understand what the concept of value is. Shazbot’s response was that my concept of value isn’t standard but that the dispute between us amounts to a semantic dispute.
Personally, I think there’s a deep issue here, one worth getting clearer on.
So let me first say that I disagree with Shazbot’s view that the dispute reduces to semantics (tho I want to recognize and applaud the effort to resolve a titchy issue between James and I!) I fully grant that James’s conception of value is correct, namely, that value is subjectively determined. My criticism – one I’ve made repeatedly on this site in exactly similar situations – is actually quite different than both James or Shazbot believe it is.
Let’s get technical. Let’s define value in terms of subjectively determined utility expectations and/or gains which justify individual actions. A person would then be justified in taking action X (exchanging chickens for medical services or cash for cars, whatever) if they believed that their subjectively determined utility is increased by doing so. Given that, it follows that since only individuals can determine whether they derive more utility from trading chickens or cash for medical services or a car, value is subjectively determined.
Fair enough. (Or … I hope it is. If the above’s not right then we need to hash it out, but I think even it isn’t the rest of the post would stand independently.) Given all that, can value be objectively measured? Well, presumably not. Values are inherently subjective and therefore (logically!!) inaccessible to observation or measurement. They exist as real, but completely internal properties of individuals.
But it seems to me that that conclusion leads to all sorts of problems. I’ll discuss two of them, arguments that people who’ve participated in some of the discussions I’ve had with others here might recognize.
The first problem I see is that a theory of political economy based solely on the concept of subjectively determined value cannot be objectively justified in terms of total utility since there is no way of knowing whether subjective utility (and therefore total utility) is actually increased (or maximized or optimized) wrt any particular policy or economic arrangement. And the reason, as I said, is that hypothesis value is not objectively observable or measurable.
This type of objection, however, strikes me as having a pretty obvious response: in a system which maximizes free and voluntary exchanges individual’s subjectively determined utility will be maximized as a matter of course since people will only engage in transactions or activities which they believe (rationally, let’s suppose) maximize their utility. And a system in which people actually can act on their own utility functions will be one which permits them – and therefore society as a whole – to maximize utility. So, by permitting people an option between A and B rather than only B, a person who freely chooses A is effectively revealing their preference for A, and is therefore maximizing their utility (relative to the two options, of course).
Now, here’s where I start having trouble. As I understand the concept, what I just described would be an instance of a “revealed preference” (or close enough…), where the consumptive act of exchanging X for A rather than B reveals individual preference rankings and relative value assignments. One thing to mention is that a revealed preference is observable and measurable (that’s the purpose of introducing and using the concept, in fact). Another is that by revealing a preference a consumer is actually demonstrating their value assignment (relative, to be sure) between a range of options as well as the marginal limit of their value assessments since there exists a price at which the exchange isn’t worth engaging in (null or negative utility).
But rather than push that line (which I think is interesting enough) I want to focus on another, more philosophically oriented, problem. It seems to me that a theory of political economy based on subjective utility functions cannot (non-circularly) be justified without introducing an objective and independent criterion. So consider: certainly it’s true that institutional policies which maximize the expression of preferences will also maximize subjectively determined individual utility (given some pretty non-controversial background claims). But that’s just the expression of a tautology since revealed preferences are defined in terms of subjective utility functions. What isn’t justified – and I don’t think can be justified without invoking objective evidence or some type of objective standard – is that maximizing the expression of subjective utility functions leads to maximizing total utility relative to competing policies or theories of or political economy. To justify that claim would require identifying objective metrics according to which the argument can proceed, something in addition to but independent from subjectively determined utility functions.
And that’s my complaint. In total. Personally speaking, I don’t think it’s a trivial one. In fact, it seems to me that libertarians invoke objective standards all the time to justify their views, and in particular, objective measurements of value. Hanley and I went ’round the maypole some while ago regarding his view that the middle class is better off now than it was in the 1950’s. My question then was: Better off according to what metric if not an objective determination of utility?
So, bringing it all back home, how does this relate to the original comment which Hanley took exception to? A common libertarian objection to the act of voting is that it’s irrational. A common libertarian response to value is that it’s subjectively determined. On the one hand, the rationality of doing X is an objective property which applies to individuals. On the other hand, the rationality of doing X is a utility function subjectively determined by individuals. That’s the inconsistency (one of them, anyway) I’m trying to point out.
theory of political economy based solely on the concept of subjectively determined value cannot be objectively justified in terms of total utility since there is no way of knowing whether subjective utility (and therefore total utility) is actually increased (or maximized or optimized) wrt any particular policy or economic arrangement.
Exactly. But if the theory of subjective value is wrong, it cannot not wrong on the basis that it leads to uncomfortable results. What you’re implicitly suggesting here is that we work backwards from what we need (or at least would really like to have) to find a theory that produces it. But as we all know, reasoning backwards from conclusions is wrong. What we end up with then is a false theory of objective value that allows to believe we are accurately measuring the total utility gains from some policy, but which actually produces false measurements.
One thing to mention is that a revealed preference is observable and measurable…Another is that by revealing a preference a consumer is actually demonstrating their value assignment (relative, to be sure) between a range of options as well as the marginal limit of their value assessments since there exists a price at which the exchange isn’t worth engaging in (null or negative utility).
Yes, but. If I buy B rather than A, you can say I value B more than A. But that’s very imprecise. You can look at how much I paid for B, but all you can say from that is that I value B at least that much. I might value it a whole lot more (and be glad I got a great deal). So how would you figure out with anything like precision how much I really value it? You’d have to figure out at what price I withdraw from purchasing it. There are means to do that (essentially, auctions), but you’re not going to be able to run auctions on everything for every consumer and come out with a good objective calculation of values.
What isn’t justified – and I don’t think can be justified without invoking objective evidence or some type of objective standard – is that maximizing the expression of subjective utility functions leads to maximizing total utility relative to competing policies or theories of or political economy. To justify that claim would require identifying objective metrics according to which the argument can proceed, something in addition to but independent from subjectively determined utility functions.
Yes, but it’s a problem more for you than for me. The fact that people are choosing their preferred outcomes means they’re at least enhancing their own utility. That is, we at least know that their utility is enhanced from T(prior to purchase) to T(after purchase). But the other policy, that conceivably might do better or might do worse at improving total utility…you don’t really have anything to measure that against. So this problem just doesn’t do you any damn good.
And finally, you really need to stop invoking libertarians in conjunction with subjective utility theory, as though it’s uniquely their baby. I call that a dishonest act because you’re trying to prejudice an idea you don’t like by linking it with a disliked group. A little guilt by association. But of course it’s not a uniquely libertarian idea by any means.Report
But the other policy, that conceivably might do better or might do worse at improving total utility…you don’t really have anything to measure that against.
You’re missing the point. It’s not another policy, it’s another metric. One conceptually independent from the concept of subjective utility. If the theory is empirically defensible, it requires an objective measure of total utility. At least insofar as the theory is supposed to be justified in terms of maximizing total utility (which libertarian theories are generally justified by).
If the justification isn’t total utility, then it’s based on a presumption that maximizing choice in the market place is an ultimate good in and of itself. That’s fine too, but personally I don’t see how that argument can go thru without invoking a conception of an objective good (ie. an objective utility calculus) either.
And finally, you really need to stop invoking libertarians in conjunction with subjective utility theory, as though it’s uniquely their baby. I call that a dishonest act because you’re trying to prejudice an idea you don’t like by linking it with a disliked group.
I’m talking to libertarians on this site, in particular, you James. So it seems entirely appropriate to focus on the group of people who advocate those views within the community I’m speaking to. Cuz I’m actually speaking to you, James, and not the readers of this blog or adherents of views that exist only in conceptual space. For the record, I actually can talk to abstract individuals, but I prefer to talk to real humans.
And where do you get the idea that I dislike libertarians as a group? I don’t. I think there wrong on precisely the terms I’ve taken great pains to identify in comments on this blog. On only those terms: no more, no less.Report
Well, then where do you get the idea that a common libertarian objection to the act of voting is that it’s irrational?
There is a common libertarian objection to strict majoritarianism, and a few who consider themselves (for whatever reason*) principled non-voters. And I suppose there’s this book**
But there’s an entire Libertarian Party that probably doesn’t think that voting is irrational. They may think voting is rigged, and they’re awful at actually getting votes, but they *are* in the game, to a greater or lesser extent.
*speaking of reason, they’ll always throw down an election post that lists whomever everyone’s voting for, but have a couple of people who have chosen not to decide and thus have made their chose.
**which was not a bad book, but I can no longer think of the author without thinking of the clone mini-me thing.Report
Well, then where do you get the idea that a common libertarian objection to the act of voting is that it’s irrational?
Because lots of libertarians at this site argue that it is without any liberal support for the view. But that’s sort of beside the point. The post isn’t meant to be understood as team X against team Y. (Tho I realize that that claim can’t be established without going meta.) And there’s this: there’s an ambiguity in the term “irrational”, one which is actually the focus of the post I wrote.Report
Well, then where do you get the idea that a common libertarian objection to the act of voting is that it’s irrational?
I think that that comes from a particular corner of “economist jokes”. Here’s one: if an economist sees another economist at the voting booth, they say, “I won’t tell if you won’t tell.”
I am willing to say that Libertarians have a higher percentage of folks who do stuff like “tell economist jokes” than the two “real” parties do.Report
Libertarianism has absolutely nothing to do with my understanding of subjective value.
Since you can’t seem to actually hear that, regardless of how often it’s repeated, I don’t think there’s any point in continuing to talk with you.Report
Nothing” seems a little strong, no?
It is easier to justify libertarianism if there is no objective account of value, it seems to me.
If value is just satisfied preferences, a world where everyone is free to act as they prefer to act, as long as they’re not violating anyone else’s freedom to act as they prefer, (it seems, at least) that you get a world with a maximum of value. Thus value is maximized by libertarianism.
That’s not to say you can’t tru to justify libertarianism in different ways. Nor is to say that this justification is something you have to agree with if you believe value is subjective.
But all that said, I think I agree with Sillwater that a lot of libertarians do accept the subjectivism that you do.
Conversely, if there is some objective value in the world that is distinct from what people prefer, the mere fact that a free (market or not) transaction gives everyone what they prefer, doesn’t mean that value has been maximized. This leaves open the possibility that a more free market society is a society with less value in it.
Again, I agree that you can believe in libertarianism without subjectivism about value or vice versa and not contradict yourself logically, but it seems to me that libertarians may pull some arguments and some normatiev intuitions from subjectivism about value.Report
The theory of subjective value was developed independently of libertarianism, and is a mainstay of economics even among liberal economists.. I pointed this out the other day. Apparently it’s an inconvenient truth, so it’s best ignored. (I stopped reading after that first clause of yours because, as with my response to Stillwater, if that’s a position you all want to stand on, there’s no value in further discussion with you.)Report
Rude.Report
“The theory of subjective value was developed independently of libertarianism…”
That is a sort of half-truth, IMO. It was developed by the Austrians, and it is at the core of their philosophy. It is also used to attack a certain kind of argument for socialism from exploitation.
The position is widely held by economists who aren’t Austrians, but this may very well be one of the failures of contemporary economists, of the sort that Sen has been pointing out for a long time.
We agree (if you’d read my comment, you’d see this) that the two views can be logically extracted, but they tend to overlap in the minds of a lot of believers.Report
value neutrality seems larger than libertarianism. It seems that liberalism as a whole is all about value neutrality (with subjectivism just being one of the ways to get there).Report
we at least know that their utility is enhanced from T(prior to purchase) to T(after purchase)
We know that? If all we know is that they made the purchase, it seems to me that all we know is that at time P(at purchase), they expected their utility to be enhanced by the purchase (or from time T(prior to purchase) to T(after purchase) if you prefer). We don’t know what effect any given purchase (or whatever state of affairs results therefrom, which itself changes over time) has on utility over time.
..Which I guess is to say that
The fact that people are choosing their preferred outcomes means they’re at least enhancing their own utility.
…is false, because the premise is false. People aren’t choosing their preferred outcomes; rather they’re trying to. They’re choosing their actions. No?
I’m not sure what the implications of this are for this debate or others, but that much does seem clear enough to me.Report
I agree with this Michael. I think it cuts right down to the problem I see as well. One thing I’ll say about it given your comment is that subjectively determined utility functions often appear as a fancy way of saying a much simpler word expressing a much simpler concept: desire. How much of the theory would be lost by using “desire” in place of the fancier “subjectively determined utility functions”? Surely it’s true that different people desire different objects in differing degrees and that desire is purely and inarguably subjective. It’s also true, however, that people desire objects which they concede aren’t good for them (negative utility) or for reasons that aren’t clear to them (irrationality).Report
On the one hand, the rationality of doing X is an objective property which applies to individuals
Irrational may be an objective property, but it is the objective property of whether someone does something which increases his own subjective utility function. Presumably if we could measure brain states* on an MRI, we would be able to point to certain correlations of brain states and actions.
*Spare me the physicalsm debate. I’m taking a short cutReport
Yeah, I agree with that. It’s the definition (+/-) of instrumental rationality.
In contradistinction, there is ends rationality. Presumably total utility? Am I wrong about that? I’m not exactly sure how that circle gets squared in the theory we’re discussing. Which is why I wrote the post, actually.Report
The argument is that voting is instrumentally irrational (at least if what you care about is effecting political outcomes)Report
But in what grounds? An objective measure, yes? Presumably, a person makes the determination that the act of voting is justified based on a subjectively determined utility calculus.
So if utility is determined by subjective determined utility functions, then why should objective measures matter at all?Report
Basically, the idea is that in order for voting to be instrumentally rational, we would have to value the outcome so highly (as the marginal effect of one vote is so smal) that it is implausible that we actually subjectively value that outcome so much. If we actually did, a whole bunch of our other actions would turn out irrational as we seemingly forgo so many other things that would have a better chance of effecting political outcomes. In fact, we would be way more involved and informed than we actually currently are. We could again work around this by saying that the activities we forgo non-voting political activity for are even more valued than political activity (adjusted for efficacy of activity). But once we have done all the work arounds, we find that we have basically inflated the subjective utility of everything and we are back at the expected subjected utility of the activities we forgo in order to vote being more than the expected subjective utility attached to the political outcome that we get for voting.Report
I’m a bit confused Still,
Maybe you can simplify for me.
Am I right that the following is your position? Even if we satisfy preferences, we aren’t maximizing utility, because utility is objective and preferences are subjective?
Is the debate here between “Preference Utilitarianism” and some kind of more objective kind of utilitarianism, like “Hedonistic (quality-of-experience-based) Utilitarianism” or Moorean “Ideal Utilitarianism”?
Here’s Preference Utilitarianism defined:
“A related position rests on the claim that what is good is desire satisfaction or the fulfillment of preferences; and what is bad is the frustration of desires or preferences. What is desired or preferred is usually not a sensation but is, rather, a state of affairs, such as having a friend or accomplishing a goal. If a person desires or prefers to have true friends and true accomplishments and not to be deluded, then hooking this person up to the experience machine need not maximize desire satisfaction. Utilitarians who adopt this theory of value can then claim that an agent morally ought to do an act if and only if that act maximizes desire satisfaction or preference fulfillment, regardless of whether the act causes sensations of pleasure. This position is usually described as preference utilitarianism.”
http://plato.stanford.edu/entries/consequentialism/#WhaGooHedVsPluConReport
It seems to me that your claim is something like the common criticism against Preference Utilitarianism, as stated here:
“Preference utilitarianism is often criticized on the grounds that some preferences are misinformed, crazy, horrendous, or trivial. I might prefer to drink the liquid in a glass because I think that it is beer, though it really is strong acid. Or I might prefer to die merely because I am clinically depressed. Or I might prefer to torture children. Or I might prefer to spend my life learning to write as small as possible. In all such cases, opponents of preference utilitarianism can deny that what I prefer is really good. Preference utilitarians can respond by limiting the preferences that make something good, such as by referring to informed desires that do not disappear after therapy (Brandt 1979). However, it is not clear that such qualifications can solve all of the problems for a preference theory of value without making the theory circular by depending on substantive assumptions about which preferences are for good things.”
http://plato.stanford.edu/entries/consequentialism/#WhaGooHedVsPluConReport
No. See a comment I’m about to post.Report
Am I right that the following is your position? Even if we satisfy preferences, we aren’t maximizing utility, because utility is objective and preferences are subjective?
It’s different than that. I’m making an epistemological argument here. It’s that even if maximizing preferences maximizes utility, there must be an observable metric by which that conclusion is measured. But if our only conception of utility is that it’s subjectively determined – inaccessibly locked in the contents of each individual’s mind, so to speak – then the theory can’t be empirically justified.Report
It’s that even if maximizing preferences maximizes utility, there must be an observable metric by which that conclusion is measured.
But there does not need to be such a metric.
Possible counter-argument: “You’re assuming that something needs to be measured; that the preference for allowing free exchange is because it necessarily leads to maximized utility. This is a mis-understanding. It has a possibility of leading to maximized utility. Other mechanisms for increasing subjective utility may also exist, and if subjective utility *could* be measured, it’s possible that in a specific case the managed scenario works better than the free exchange.
But it is simply more improbable than free exchange due to the number of actors involved, and information loss. We don’t have to measure subjective utility gains in two difference scenarios, we’re simply assessing the likelihood that two agents come to a subjective utility maximization between themselves as being more likely than two agents coming to a subjective utility maximization when a third party is involved in an adjudicator role.”
Let’s say you have two agents, each with a random pile of colored jelly beans, and each with a preference for which colors that they prefer. You allow them to exchange the jelly beans on any basis (single bean for single bean, single bean of one color for multiple beans of another, etc.) At the end, they both still have two piles of jelly beans, but they’re no longer randomly distributed as both agents have expressed their preference by trading beans for beans.
Now let’s say you have three agents, two with a random pile of colored jelly beans, and each with a preference for which colors that they prefer. The third agent does not *know* the ordered preferences of the other two agents, but they have established rules of trading jelly beans. Perhaps there are limits on the number of beans that can be traded for a single bean, or specific other rules (red/raspberry beans can’t be traded). At the end of the trading cycle, again you have two piles of beans that are no longer randomized.
In which scenario is it more likely that both agents will be pleased with the outcome of the trading?
The free-exchange-leaning person will say #1. The managed-economy-leaning person will say #2.
I say, “It depends on what the piles of beans look like in the end, for each iteration of bean trading”, because the funny thing about people is that they can make all sorts of individual decisions that they liked while doing the trading but the end result is something that they decide is unfair. Which is kinda why I’m neither a free-exchange guy or a managed-economy guy; I think they’re both right, and they’re both wrong.
Sorry about closing the thread on you, Stillwater. Hopefully I won’t have to do that again anytime soon.Report
It’s possible, I suppose, for the claim “voluntary exchanges between individuals maximize total utility” to be a priori true. It’s certainly possible that some people make that argument. It’d be strange, tho, for a theory of economics to accept rationalism to that extent. And stranger yet for that views adherents to admit to that level of rationalism.Report
That’s not (necessarily) the claim, though, Stillwater.
The claim is that you can arrange different mechanisms for increasing subjective utility, and that the mechanism of private exchange is most *likely* to increase total utility because you’re not putting in a third agent.
It’s an information problem. Two agents (presumably) know themselves better than an outside agent knows either of them. By adding the outside agent, you’re increasing the likelihood that the final outcome will have inefficiency due to information loss.
You don’t need to be able to compare the two outcomes directly to make that statement. It’s like that old saw about a feminist bank teller.
Given three agents, what is the likelihood that the third agent knows the other two agent’s utility functions well enough such that the third agent is contributing to the total subjective utility?Report
I’m thinking mainly that you can move far enough on a subjective measure that it approaches asymptotically to objective. But strictly speaking, a subjective measure that almost everyone agrees on is not ipso fact an objective measure, though it may look like one at a glance.Report
You’ll have to elaborate on that for me to understand, K. It sounds like an interesting idea.Report
everybody agrees ‘thou shall not kill’. But it is really everybody, or just an overwhelmingly majority to the 99+% that it looks like everyone, and anyone who thinks otherwise is considered a socio/psychopath?
But while a sociopathic view is sociopathic, it is still a view.
(and that’s putting aside that what some cultures would consider murder, others would not. Plus differing norms on how the State is authorized to commit violence and the final sanction)Report
Sociopaths are different. There is a deeper set of practices that involve the way people form and invoke reactive attitudes like resentment and indignation that sociopaths just do not participate in. Sociopaths are just don not feel indignation at anything. That’s why they are not even part of any moral community and not just people who happen to disagree with the vast majority of their fellow members.Report
“You’ll have to elaborate on that for me to understand, K. It sounds like an interesting idea.”
Example: “Eating horses, dogs, cats, ferrets, and other pet animals is a moral outrage.”Report
Or maybe you mean to say the following: I can always meaningfully ask if X’s subjective preferences are really good, valuable preferences that X ought to have.” This question is meaningful only because the statement “X’s preferences are really good, valuable preferences that X ought to have” is a synthetic statement, i.e. it is not true in virtue of the meanings of the words alone.
But if “Is valuable and good to X” and “Is X’s preference” mean the same (if one is the correct analysis of the meaning of the other), then the statement “X’s preferences are really good, valuable preferences” ought to be an analytic statement (i.e. true by definition like “All Bachelors are unmarried men” or “A = A.”)
Of course, this is the Open Question argument, which works to invalidate, not just any form of preference utilitarianism, but all forms of utilitarianism and lots of others ethical theories besides.Report