PLSC 118 - Lecture 5 - Classical Utilitarianism and Distributive Justice

Chapter 1. The Measurement of Utility [00:00:00]

Professor Ian Shapiro: Okay, this morning we're going to carry on talking about Jeremy Bentham and classical utilitarianism. And I'm going to begin by making a few points about the measurement of utility, which we bumped into in a glancing kind of way last time, but we're going to dig into it a little bit. And then we're going to move from that into talking about utility and distribution in classical utilitarianism; how we should think about the measurement of utility across the whole society and what implications Bentham's argument has for that, and I think you'll start to see why classical utilitarianism became such an ideologically powerful doctrine in the eighteenth and nineteenth centuries.

So just briefly to recap, we talked last time about Bentham's principle being maximized, "the greatest happiness of the greatest number." The idea being that if you think of, in this case, a very simple two-person society, and you think of that as the status quo, A has that much utility, B has that much utility. Anything on this side of the status quo would be an improvement for society. The greatest happiness of the greatest number will have been increased.

Now, that's all very abstract, and by way of trying to make it somewhat more concrete, let's notice two features of utility measurement. The first, as I said to you last time, as far as Bentham is concerned, this is a doctrine of what I called objective egoism; that people are self-interested and behave self-interestedly, but that we can figure out what's likely to motive them regardless of their own interpretation of their actions or behavior. We have our interpretation; remember, Bentham says the rare case — as with the physiology of the human body, so with the anatomy and physiology of the human mind, it's the rare case that you get it right about yourself. And it's the objective scientific calculus that's going to tell us what maximizes people's utility.

Now, you might say, "Well, how is that actually going to work?" So there are two steps here. The first one is that he thinks all utility is quantifiable. I went through that last time, but the piece I didn't mention is that it follows from that that utility is reducible to a single index, and in this case Bentham's thinking of money. Money is going to be the measure of utility in his scheme, and that means that we could think of these units of utility as having a kind of dollar value. So anytime you think this doctrine is crude or extreme, remember my point that this is a guy who takes every thought to the logical extreme.

But if so you get one, let's just for simplicity say, say one Standard International Util costs a dollar. And let's suppose you experience two Standard International Utils of pain from coming to class. Then I could make you indifferent between coming to class and not coming to class by paying you two dollars. It could get you to come to class if I paid three dollars, and I would not get you to come to class if I paid a dollar, right? And so that's the second point. In the first instance we say that utility is quantifiable and expressible through money, but then related to that, and as indicated in the example I just gave you, we can work with a doctrine of revealed preference. We can vary the price that we charge admission for the course. So let's say we charge — let's imagine there are three of you and one of you experiences two utils of pain from coming to class, one experiences three utils of pain, and one experiences two utils of pleasure. There's one perverse student in the audience who actually likes coming to the class. So then we would find that if we paid a dollar, one of you would come. If we increase it to two-fifty, two of you would come, and so we could vary the price to get information about your utility.

And we could even influence your behavior without actually changing your preferences, and that's a very important distinction to make. Your enjoyment from coming or not coming to class wouldn't change, but your behavior would change if we varied the price; so that we can influence your behavior by manipulating the incentives without regard to what your underlying preferences are, and we could allow them actually to stay the same. You'd rather be at home asleep, but if the price is high enough you'll come anyway.

Chapter 2. Classical Utilitarianism, Distributive Justice and Diminishing Marginal Utility [00:05:57]

Okay, so that's all well and good at the level of thinking about one individual's behavior, but what about thinking about society in more general terms? When we talk about utilitarianism in Bentham's system, classical utilitarianism, we see that he operates with these numbers that attach to specific actions or policies and that we can make comparisons across individuals. So to put this in the jargon of economists, Bentham allows interpersonal comparisons of utility. Bentham allows interpersonal comparisons of utility. We can say that if you take one unit of utility from one person and give it to another person their utility will go up and the first person's utility is going to go down. Okay, so it's a doctrine of interpersonal comparisons of utility.

And for those of you who are mathematicians here it might also be worth noting that Bentham operates with cardinal scales. These are additive things. You can actually think about these as sort of lumps of pleasure or pain experience that are moved around across people and can be added and subtracted. And so I put up just here to sort of — so you can think your way through this doctrine. If you imagine a status quo, a perfectly egalitarian world in which each person has six units of utility, you can start asking yourself, "Well, let's imagine if we could redistribute things." What would that mean as far as Bentham's doctrine is concerned?

What I've given in this first column as a potential departure from the status quo is the utility monster example we talked about last time. If it turns out Leonid has a vastly superior capacity to experience pleasure than anybody else, then we could get a huge increase in total utility by taking a lot from B and C and giving it to Leonid, so that we would say "allow," right? Or we could think of this change from the status quo — we go to a more inegalitarian society and, again, the greatest happiness of the greatest number has increased. We have a world here where there are eighteen utils and a world here where there are nineteen utils.

Or think about this case, we might think of this as a kind of schematization of the Eichmann problem. If the utility that the Aryans gain from practicing genocide and ethnic cleansing against the Jews exceed the utilities that the Jews lose, there would be no reason under Bentham's doctrine not to do it.

Okay, now there's a certain ambiguity in the phrase, "Maximize the greatest happiness of the greatest number," which Bentham never finally resolves. The ambiguity is whether he's saying just maximize the total, so here the total's bigger than eighteen, here the total's bigger than eighteen, here the total's bigger than eighteen, so it's obviously the case that it's preferable, on Bentham's scheme, to the status quo.

Or is he perhaps saying maximize the utility of the majority, the greatest happiness of the greatest number, the greatest number simply meaning the majority? But in that second interpretation you could still get highly inegalitarian distributions being judged superior to the status quo, because if you imagine going from here to here we've got a majority here experiencing twelve utils of pleasure, and here we have a majority of two experiencing seventeen potentially utils of pleasure. So there is some ambiguity there as to just what Bentham meant, but most of the time he is taken as having meant just the crude statement, "Maximize the total amount of utility in the society." And so that nuance between whether we're saying the greatest number means a majority or just the total amount is not something that will detain us any further.

Now, you could say, "Okay, so far so good, but isn't all of this a little counterintuitive?" After all, if you compare — let's focus on the difference between the status quo and distribution IV here. These people might be on the verge of starvation. Surely giving them a unit of utility is going to be much more enhancing to their happiness than giving A a unit of utility. Anyone know what the principle behind that idea is? Anyone want to take it? How many of you have done ECON 101, the first econ course? Yeah, so what is the principle that would tell you if you have no food and I give you a loaf of bread, your utility goes up a lot more than if I have ten loaves of bread and I give you a loaf of bread. Somebody? Okay.

Student: Diminishing marginal utility.

Professor Ian Shapiro: Diminishing marginal utility, the principle of diminishing marginal utility of all good things. And this is the idea just encapsulated, to make it a little bit more dramatic: if you don't have a car and somebody gives you a Porsche Turbo, your utility's going to go up a huge amount, right? But if you already have a Porsche Turbo and somebody gives you a second one, you're going to get less new utility from the second Porsche than you had from the first. And if somebody gives you a third one you're going to have less utility, less new utility from the third one that you had from the second. It's not that you won't get any new, but you'll get less.

And the principle of diminishing marginal utility says that this line will get flatter, and flatter, and flatter toward infinity. You'll always get more utility from a new increment of the same good, but it'll be less new utility than you got from the previous increment of that same good, okay? That's the concept of diminishing marginal utility. The new utility you get diminishes at the margin. Each new Porsche is less valuable to you than the previous Porsche. Now, is that plausible? Anyone think there's a problem with that idea? Yeah?

Student: The idea with shoes. If you're given one shoe you're going to get absolutely no utility, but if you're given two shoes, a right and a left, then maybe you'll get more utility?

Professor Ian Shapiro: Okay, so shoes. If we just kept giving you lots of right shoes, there'd be a problem.

Student: Right.

Professor Ian Shapiro: Okay, so I think Bentham would have to say it would have to be pairs of shoes, right?

Student: Yeah, I guess.

Professor Ian Shapiro: Okay, but that's a great example to start us off on this. What else? Anything else anyone might find problematic? Yeah, over here.

Student: Well, it just seems that if we're going by diminishing margin utility that if you had everyone literally dirt poor and always starving, if you give them just a little bit of something their happiness would increase so much more because they got that much little, but people would still be living in misery technically.

Professor Ian Shapiro: Just explain that a little more.

Student: Well, marginal utility is if you had a little bit of something for the first time your happiness increases so much more because the first time. So if you were to give people very little food, or anything at all, and then you suddenly gave them a little bit they would get really, really happy about it. But by this then also if they're also very wealthy and they got something more they wouldn't really be happy. So it'd be more beneficial to the utility if they only got a little bit so they would be very, very happy.

Professor Ian Shapiro: Okay. That's a very sophisticated observation. I'm just going to put it one side and come back to it in about ten minutes when I start talking about redistribution. Okay. But anything else about — if we're still focusing on one individual, anything else that might be problematic with this notion of diminishing marginal utility? Over here?

Student: Well, let's say you're C. Just because you're rich doesn't mean you don't want to be more rich, and just because you have a certain amount of money doesn't mean more money isn't going to make you equally as happy as it did before.

Professor Ian Shapiro: Okay, that's true, but why is it problematic?

Student: I don't know.

Professor Ian Shapiro: Well, I think you hit on something really important. There are a lot of people who think that the principle of diminishing marginal utility means that money is less important to people as they have more of it. After we said the principle of diminishing marginal utility of all good things, right? Money is a way of purchasing good things, so your example might be thought to suggest that this implies the more money you have the less important money is to you, okay? So you're right, but notice what that means. Does it mean that rich people will care less about money? It's a tricky question because the first impulse is to say, "Yes, they'll care less about money," but the answer is no. Why is the answer no? Yeah?

Student: They just need more money to get the same amount of happiness.

Professor Ian Shapiro: Exactly. They need more money to get the same amount of happiness precisely because of the principle of diminishing marginal utility. So you got it exactly right to see that money creates some problematic examples for the principle of diminishing marginal utility. But the thing that follows from it is that, for Donald Trump to get more utility, you have to give him a huge amount of new money just for him to get the same amount of new utility as somebody who only has ten thousand dollars, right? So the way to think about the desire for money it's a bit like sort of a heroin addict needs more, and more, and more new heroin to get the same hit, right? So the more money you have, actually the more money you will want in order to get the next marginal increment of utility. So we should expect rich people to be greedy by this theory, not to become more and more indifferent to money. Very important assumption and a lot of people get that wrong when they think about the principle of diminishing marginal utility.

Are there any other examples of this doctrine that might make it seem problematic? Yeah, over there.

Student: Well, if I had a second Porsche Turbo I would be just really reckless with it and I could do whatever I want. I wouldn't have to protect the first Porsche Turbo as much.

Professor Ian Shapiro: Yeah?

Student: I mean it's like there's more you can do with it, right?

Professor Ian Shapiro: Yeah, so why is that a problem?

Student: Well, then wouldn't the second car — I mean, like, say if you have a little bit and you're given a little bit your utility goes up, but you really want to protect that little bit, but when you get more maybe it encourages you to save money, to not spend more.

Professor Ian Shapiro: So you wouldn't want the second one?

Student: What?

Professor Ian Shapiro: Are you saying you wouldn't want the second one?

Student: Well, why wouldn't I want the second one?

Professor Ian Shapiro: If you had one and I said, "I'll give you my one, it's right out there," you wouldn't want it?

Student: It's not that I wouldn't want it.

Professor Ian Shapiro: You wouldn't be like Jay Leno, who — how many cars does Jay Leno have?

Student: Too many. It's not that I wouldn't want it, but maybe the utility for the second one in some cases would be more than the utility for the first one so the curve would be thrown off.

Professor Ian Shapiro: Because?

Student: Because you want to protect that first one, so, I mean, so you don't lose what little you have.

Professor Ian Shapiro: Okay, so it's a possibility. Any other examples of where this becomes problematic? I mean, think about beer. One beer increases your utility a lot. The next, and the next, and the fourteenth, isn't it going to at some — you know, or taking an aspirin, isn't it going to, you know? No?

Student: What about other values like integrity? If you have a little bit of integrity and you get some more, but if you have a lot of integrity, a little bit more is still worth an equal amount.

Professor Ian Shapiro: So integrity is a great example because once you start putting values like that out there it, I think, threatens the idea that it's all reducible to a single index, right? Because you can't — having a little bit of integrity is sort of like being a little bit pregnant, right? Once Eliot Spitzer's integrity is blown it's not like there's some — it's a binary good. You either have it or you don't, right? People either think he's either a hypocrite or he's not, it' a binary thing. Maybe some people are somewhat hypocritical, but it seems like there's a threshold there, one side or the other. So there might be some goods like integrity that are not easily capture-able in this logic. We should put that out there, but yeah, over here?

Student: What about health? It's not quite binary because you can be in medium health, but I think it would be pretty useful to be healthy and then super healthy, ad infinitum.

Professor Ian Shapiro: Health. The thing about health it's tricky. Actually less so in our day than Bentham's, it's tricky to think about redistributing health, right? Although you'll see we will come up against some pretty bizarre cases. If some people are sighted and some people are blind and you could do eye transplants, should we be transplanting from the sighted to the blind? Arguably the blind person would gain more utility from getting one eye than the sighted person would lose from losing one eye, so shouldn't we do that? So that can also give you some ways of proceeding that would make you queasy, right, if you allow the principle of diminishing marginal utility.

What about the examples I threw out there, beer and aspirins? They're a bit like the sort of left shoe examples, right? I don't think those are actually deep problems for Bentham's theory because I think what he would say is, "Well, you'd drink beer and at some point you would sell the beer rather than make yourself paralytically drunk and feel terrible." You'd sell the beer and use that to buy some other good that would give you increasing utility at a diminishing marginal rate. So the main thing is that the fungibility of utility and its expressibility in terms of money, although as was pointed out here, when we think about the diminishing marginal utility even of money, we shouldn't think that that makes you care less about money the richer you get; rather it will make you care more about money the richer that you get.

Okay, now, here's a historical statement about the principle of diminishing marginal utility. Every serious economist since the eighteenth century has assumed that the principle of diminishing marginal utility is true, including Jeremy Bentham. You can't do economics without assuming that the principle of diminishing marginal utility is true. And I think if you threw out some of these problematic instances like integrity, I think that what Bentham would have said, or what any economist would have said, "Well, yes, there are some things that are not capture-able easily, or easily captured by this idea, but if you want to get it right, if you want to see how people are going to behave, if you want to get it right, it's a better assumption than any of the competing assumptions you could make. It's going to get you closer to the truth more of the time than not assuming the principle of diminishing marginal utility is true." So Bentham would have probably said that, I think, if questioned or if somebody had probed with some of these counter examples. So it's the best assumption you can make given that you've got to assume something.

Chapter 3. Diminishing Marginal Utility, Practical and Absolute Equality [00:25:57]

But now, and now I want to come back to the sophisticated point that was made in the middle at the back there a few minutes ago, when you start to think about the utility that people at the bottom of the social order derive from a particular good, versus the utility that the people at the top of the social order derive from some particular good, because in Bentham's scheme, remember, we are allowing comparisons across individuals. Let's suppose a two-person society, again, and let's suppose it consists of Donald Trump — well, it can be a multi-person society but we're just going to focus on two: Donald Trump and a homeless woman living out of a left luggage locker in Grand Central Station. Actually there are no lockers at Grand Central but there are at Penn, at Penn Station, okay? And the question is, should we take a dollar from Trump and give it to the bag lady. What? Should we? Yes? No? How many think yes? Okay, yeah, almost everybody.

Why? Because by assumption with the principle of diminishing marginal utility we take the dollar from Trump up there, his loss of utility is negligible, but we give it to the woman who's starving down here, and her gain in utility is enormous from that dollar, right? So we should take the dollar from Trump. Let's assume there's no dead weight loss to the government and all of that for right now. We will just keep it simple. We should take that dollar from Trump and we should give it to the bag lady, and the greatest happiness of the greatest number will have increased, right? But then maybe we should take another dollar, shouldn't we? I mean it worked the first time, so we should take a second dollar from Trump and give it to the bag lady, and a third dollar, and a fourth dollar. When are we going to stop? When are we going to stop? Yeah?

Student: [inaudible]

Professor Ian Shapiro: Yeah, we're going to stop at the point of perfect equality, right? We're going to keep redistributing until they have the same amount. So now you should be able to start to see why classical utilitarianism was a doctrine that was thought to be profoundly radical and frightening to rich men, because it has this built-in impetus for downward redistribution. You can say well, there'll be cost, there'll be dead weight loss to the state and so on, but still the underlying logic says take it from Trump and give it to the bag lady, right? At the margin that's what you should do.

And Bentham completely saw that this was an implication of his doctrine. Now, Bentham was a fairly radical guy. He was a supporter of democracy, which was a radical thing at that time. But he wasn't as egalitarian as all that, and he wanted to temper the downward redistribution that flows from his principle, and so he makes a distinction between what he refers to as "absolute" and "practical" equality. He says,

Suppose but a commencement made, by the power of a government of any kind, in the design of establishing it (absolute equality, that's redistributing to equality), the effect would be — that, instead of every one's having an equal share in the sum of the objects of general desire — and in particular the means of subsistence, and the matter of abundance, no one would have any share of it at all. Before any division of it could be made, the whole would be destroyed; and destroyed, along with it, by those whom, as well as those for the sake of whom, the division had been ordained.

He's basically saying, if you want to reduce that to a bumper sticker, he's saying the rich will burn their crops before giving them to the poor, and that is a common argument in politics. It's the sort of reverse of trickle-down, right? Trickle-down is the notion that you allow inequality because the rich will create more wealth for everybody, right? The pie bigger for everybody, and so the greatest amount of utility is increased by allowing inequality. This is the inverse claim. Bentham's saying, "Well yes, in principle absolute equality would maximize the greatest happiness of the greatest number, but in fact if a government set out to do that, the rich would rebel." 

And this is a claim that is often made in everyday politics. So you'll destroy incentives to work, is the claim that you'll hear when we have arguments about raising taxes in the run up to the fall elections, right? In the transition to democracy in South Africa people said the white farmers will destroy their farms before turning them over to the majority. It turned out not to be true. So those examples put on the table, what sort of force does this claim have? It's really an empirical claim, and we don't really know how much the rich will tolerate before burning their crops. Presumably they'll allow some redistributive taxation, but we don't know how much, and a lot of the day-to-day argument of politics turns around how much.

So Bentham makes a distinction between absolute and practical equality, and he says, "We should redistribute to the point of practical equality, but not to the point of absolute equality because redistributing beyond practical equality has this perverse counter-trickle-down logic and that's not going to be acceptable from the standpoint of the principle of utility."

Okay, so when you allow both interpersonal comparisons of utility and you assume diminishing marginal utility, utilitarianism becomes a very radical doctrine. You can hedge it in to some extent with claims of this sort, but they are themselves controversial and you're going to get into a very messy world of macroeconomic predictions and counter-predictions about whether and when you reach this point of practical equality, or when the gains from downward redistribution are offset by the losses from the shrinking of the pie.

Chapter 4. What about Rights? [00:34:15]

Now, some of you might have said, "Well, at the beginning of this course of lectures, Shapiro said, 'Every Enlightenment thinker is committed to postulants. One is that we can have a scientific theory of politics, and the other is that individual freedom operationalizes a doctrine of rights is the most important good.' Now, having sat through these lectures on Bentham, I can see what he's saying about science. Bentham has this monomaniacal view of science. He's got his objective egoism. He can figure it all out, what will maximize social utility, and run around the world writing constitutions for people, can devise a whole public policy that's going to scientifically maximize the utility of society, but I'm not seeing a whole lot of room for rights in this doctrine. It seems to allow ethnic cleansing, even genocide. It seems to allow redistribution from one person to another, all justified on the grounds that this is maximizing the total utility of society. Well, even if it is, how does this respect individual rights?"

Am I just wrong? Is there some elementary thing I've missed here? There's not much room for rights in Bentham's doctrine. So I'm just wrong that these Enlightenment thinkers were committed to individual rights? It would be a reasonable inference from what I'm said so far. But remember, for Bentham when we try to maximize utility in the society, individual motivation is vital. This is a passage I read to you last week, but I'm just repeating it, "The great enemies of public peace are the selfish and dissocial passions — necessary as they are...Society is held together only by the sacrifices that men can be induced to make of the gratifications they demand: to obtain these sacrifices is the great difficulty, the great task of government."

He's saying you have to work with individual motivations. You can't ignore them, and I think that is the point that's behind his distinction between absolute and practical equality. The rich will burn their crops before giving them to the poor. You have to take that into account. You have to see individuals as the basic generators of utility. In another piece of Bentham's writing which I didn't have you read, but I'll just put it out there because it's where you start to see our old friend the workmanship ideal creeping by the backdoor into utilitarianism. Bentham says, "Law does not say to man,Work and I will reward you but it says: Labour, and by stopping the hand that would take them from you, I will ensure you the fruits of your labour — its natural and sufficient reward, which without me you cannot preserve. If industry creates, it is law which preserves. If at the first we owe everything to labour; at the second, and every succeeding moment, we owe everything to law."

So another way of thinking about this is, that Bentham's idea of the state is essentially regulatory. It stays the hand of somebody else who would steal your goods, but the government cannot itself create utility. Labor creates utility, and this is why I say that workmanship, that idea that we first confronted when we talked about Locke, comes into utilitarianism by the backdoor, because Bentham's going to say, "Unless you respect individual rights you're not going to be able to maximize utility for the society as a whole."

So the state is basically a regulative state, not a state that's actively involved in creating utility for individuals. It will do some redistribution to the point of practical equality, but the basic idea is that the state should be hands-off with respect to the utility creation in the society. It's industry that creates utility — labor, work — so incentives are going to be important going forward if you're going to maximize utility. So that's the way in which we see that even a classical utilitarian like Bentham is going to resist dispensing with the doctrine of individual rights.

Now, there's a problem, though, with his mode of doing this, and the problem arises because the claim that the rich will burn their crops before giving them to the poor might not be true. And even if we get to less extreme circumstances like South Africa before and after the transition, when we look at actual debates in contemporary politics in the United States, this is what we see. Ronald Reagan comes in and says, this is in 1980, "If we cut taxes, the pie will get bigger for all and they'll be actually more revenue," and so utilitarianism says do it. And the Democrats say, "No, they won't," and it's an empirical argument. And you will find, if you go back now and look at what happened during the 1980s, perfectly credible economists will line up on both sides because they cut the taxes, but, of course, eight other things happened as well that affect the macro-economy, right? And disentangling how much the tax cuts were responsible for what happened, versus how much many other things that happened were responsible, nobody really knows.

Or if you look at the current debate we watched and are watching unfold about the economic stimulus. If the economy turns around between now and November, the Democrats will probably do a lot better than if it doesn't, but the Republicans will say, "Well, it would have turned around faster if we hadn't had all this taxation." And Paul Krugman will say, "Well, it would have turned around even faster if we had had more taxation." And so a lot of the problem in debating incentives, once you get into the real world of macroeconomic policy-making, is that (a) you never have the counterfactual; you can't go and rerun history without the stimulus, right, or without the Reagan tax cuts. And (b) the sheer complexity; so many other things happened — the price of oil goes up, or the commodities collapse, or the dollar, or this, or that, or the Chinese revalue, do or don't change the value of their currency.  

So that when it gets down to it, you're never going to get a definitive answer to the question what is the point of practical equality. When have we passed the point of practical equality, to use Bentham's terminology? Are we close to it? Have we gone by it? Are we nowhere near it? There have been periods in our history when we've had top marginal tax rates of 90 percent, right? Reagan thought a top marginal tax rate of 40 percent was beyond the point of practical equality. You're never going to get a definitive resolution of those questions.

But if we think back to what the aspiration of the early Enlightenment was, it was certainty. To use the example, remember, I read to you from Hobbes, from his Epistle Dedicatory to his Six Lessons to the Professors of Mathematics; he said, "For the things we don't make, we can't know we can only guess about the causes," right? Well, here we're guessing about the causes. We don't really know and there will be — the people who want either policy will be able to find a plausible set of experts to defend their view. So you're getting to this very messy world of macroeconomic prediction, if you want to put some limits on the radical edge of classical utilitarianism. And as a matter of history, that's not how it went.

As a matter of history, how it went was to rethink the analytical structure of utilitarianism in a way that completely defanged its radical redistributive edge without any reference to these messy macroeconomic considerations. And just how that happened in the transition from classical to what we're going to all neoclassical utilitarianism is a subject with which I will begin on Wednesday. See you then.

[end of transcript]