June 30, 2004

On the Maximization of Social Welfare

Via Elephant's Trunk, Max Sawicky lightens his Karmic burden by encountering a law professor who makes a mistake--a mistake that is mademuch too-often made by those who have taken too few courses in welfare economics, and also attended too many seminars sponsored by the Olin Foundation:

Elephantstrunk: Max Sawicky called a Prof. Bainbridge on a rather tasteless criticism of social democracy. Prof Bainbridge responds with:

I find the GNP data useful because I start with an assumption that social wealth maximization is the appropriate goal of public policy. This position was developed by Judge Richard Posner*, among other founders of law and economics, who argue that society should look to wealth maximization rather than utility maximization. The difficulty of making interpersonal comparisons of utilities, among other drawbacks of pure utility analysis, requires that we monetize preferences in order to say anything meaningful.

Why Prof. Bainbridge thinks that he has a right to a simple theory I don’t know. However his wonderful frankness makes the decisions made clear. He, and apparently others, have chosen a theory that they can work with over one that has meaning, they would rather say true things that may have no meaning than things that might not be true but really do have meaning...

To recap, Bainbridge says (i) we can't pursue the utilitarian goal of maximizing social welfare (ii) because we cannot compare individual utilities, so (iii) let's take our goal to be the maximization of wealth.

But Bainbridge is wrong in claiming that the goal of wealth maximization allows one to dodge the problems of weighting individuals' preferences. It resolves them in a particular way, and in a way that people like Bainbridge appear to have difficulty seeing.

Let's suppose that we accept Bainbridge's suggestion that we shape public policies with an eye toward maximizing wealth. What then are we doing?

The first thing we are doing is trying to make the sum of the wealth of all the members of society as big as possible. Letting Ws stand for individual members' wealth, we are then maximizing, for members 1, 2, ... n of society:

Let's consider some quantitative indicator x of legal policy--damages for negligence or rights of shareholders or something else. By varying x and looking across societies with different values of x, we also vary each individual's wealth as the different legal rules interact with private decisions to produce different outcomes. Let's use standard derivative notation: dWi/dx is the ratio of the change in individual i's wealth per unit change in x. If infinitesimal increases in x increase total wealth, we move in that direction; if infinitesimal increases in x decrease total wealth, we move in the other direction. We stop--we decide on what is the appropriate value of x, and so maximize social wealth, when infinitesimal changes in x have no effect on total wealth. Thus we can think of the problem of choosing legal rules to maximize social wealth as solving this master equation (plus appropriate second-order conditions):

But what are we doing, from a utilitarian individual preference-based perspective, when we solve the equation above? Let's look at each individual term: dWi/dx. We can write each of these terms--how fast wealth changes for individual i as the legal rules x change--as a product of two terms. The first term is how fast wealth changes as the individual's utility changes. The second term is how fast the individual's utility changes as the legal rules x change.

We can then rewrite this as (a) how fast the individual's utility changes as the legal rules x change divided by (b) how fast the individual's utility changes as the individual's wealth changes. The bottom term has a name: the marginal utility of wealth:

The individual terms in our master equation that we must solve to maximize social wealth can thus be thought of as the effect of policy on individual utility times an individual specific weight, which we will call wi.

What do these wi's look like? Let's consider a particular utility function, in which utility is equal to the logarithm of wealth:

This utility function captures the idea that equal proportional increments to wealth have equal effects on utility, satisfaction, individual welfare: doubling your wealth has the same psychological and welfare benefit, whether your initial wealth is $50,000 or $5,000,000. This accurately captures our intuitions about the declining marginal utility of wealth, and allows us also to come up with relatively simple formulas.

If people have this logarithmic utility function, then theirmarginal utility of wealth takes a very simple form, and the individual weights wi take a particular simple form as well: each individual's weight is simply his or her wealth:

Solving the master equation for social wealth maximization is thus the same thing as solving an analogous equation for utilitarian social welfare maximization--setting the sum of the effects of a policy change on individuals' utility to zero--only we are multiplying each individual's marginal utility change by a weight, that individual's wealth:

So social wealth maximization *is* utilitarian social welfare maximization--only with each individual's utility multiplied by his or her wealth (if utility is logarithmic; or by a factor that grows faster than wealth, if the marginal utility of wealth falls more quickly than log; or by a factor that grows slower than wealth, if the marginal utility of wealth falls less quickly than log).

Saying that you are maximizing social wealth hides the interpersonal utility comparisons that you are doing away in the background; it doesn't avoid them. And it commits you to a particular scheme for weighting individuals' utilities: the richer you are, the bigger the weight that you get in the "social welfare" function that implicitly underpins the maximization of social wealth. This, then, is the social welfare function that social wealth maximization commits you to:

Multiply everyone's utility by their wealth, and then add up all the products. This is certainly a different social welfare function than the standard utilitarian simple-sum-of-utilities. And now we see that the right question to ask is, "Is this a sensible objective, is this a sensible goal for a social welfare function?" I would answer, "No." It is hard for me at least to see how one would construct an argument that would sustain Bainbridge's position--that would answer, "Yes."

*I had always understood Posner as making a different argument than that wealth maximization was good because making interpersonal comparisons was "difficult". At least as Marty Feldstein and Richard Musgrave explained Posner's underlying position to me, it was that *common law judges* should adopt *wealth maximization* as a rule to guide their construction of common law rules. As they explained it to me, distribution was the province of the legislature and efficiency was the province of the common-law courts, and this was a sensible institutional division of labor.

Posted by DeLong at June 30, 2004 06:06 PM | TrackBack | | Other weblogs commenting on this post

You repeated the top post, FWIW.

Hope this helps.

Posted by: Alex on June 30, 2004 06:20 PM


Less taxes = more wealth :)

Posted by: david on June 30, 2004 06:31 PM



I suspect that they would immediately challenge the assumption that U = ln(W). I would also assert that x is bounded, but you can even in that case you can make the same arguement with Lagrangian.

Posted by: blank on June 30, 2004 06:35 PM


I'm just now looking at one of Sen's books ("Rationality and Freedom"?). I've just looked at his introductory, synoptic chapters, but is seems that he goes to great effort arguing points that everyone except economists has always believed (without proof). That is, some comparability of welfare is possible (E.G., a life-saving drug produces more welfare than a throwaway gag gift of the same price), and that easily-defined clear definitions of "rationality" (consistency, self-interest maximization, and one other maximization (?)) are not really very good guides for life. So more power to him, but he really seems to have brought economics back to the common sense of practically everyone before economics came along.

I run across this in many fields (also philosophy and psychology) -- the idea that well-defined, formalized ideas which are part of a system are superior to common sense, even if the well-defined ideas really seem to be wrong.

I think that a lot of the anti-ethical, value-neutralist tendency in academic life (and that's what Bainbridge's principle was) comes from the post-WWII revulsion at extremist moralizing philosophies like Marxism. (Even though Leninism is totally cynical, Marxists still make ethical critiques of others). Naziism isn't exactly moralizing, but it pretends to fill public life with vast meanings, and people like Popper Berlin, and Milton Friedman moved toward a more neutral public philosophy.

Posted by: zizka / John Emerson on June 30, 2004 07:14 PM


The Coase Theorem. Will somebody PLEASE think of the Coase Theorem?

Posted by: a on June 30, 2004 08:57 PM


It goes back a little farther to around the First World War I to Pareto, who got the idea from Walras. I always thought that the 'value-free' utility criterion was an attempt to paper over the extremely divisive debates of the 1910s through 1930s over the distribution of wealth and power in democratising societies. The efficiency argument for maximizing wealth (more is generally better than less, cet. par.) was something most people could agree on, even if they couldn't agree on who should get most of the more. A minimum criterion would be the one invented by Pareto -- if someone gets a little bit and nobody gets any less, than a case can be made that the aggregate of everbodies has more. It's a dodge, but one that seemed to work at the time.

There's an analogy to this debate in the rise of modern philosophy under Descartes, who also tried to sweep ambiguity under the table in a time of deep conflict -- that time religious rather than economic.

Posted by: knut wicksell on June 30, 2004 08:58 PM


Toulmin's "Cosmopolis" is a great book about the early modern period. Popkin's "History of Skepticism" (?Title?) is also interesting.

"The Problem of Unbelief in the Sixteenth Century: The Religion
of Rabelais" by Lucien Febvre was no fun to read, but interesting: apparently at the time of Rabelais there were almost no real atheists, but everone accused everyone else of being an atheist in different ways.

Ethical and political questions are scientifically undecidable to a large extent, but you can't quite bracket them out either. To say that there's no welfare function is to make a number of definite statements about what's tolerable, which believers in a commonsense "welfare function" could never believe.

Posted by: zizka / John Emerson on June 30, 2004 09:34 PM


Maybe the way to go about this would be to develop a general theory of what happens in the whole system, BETWEEN the events of market transactions. In other words, current standard economics would only be about ONE-HALF of total real economics. ("About" one-half, because it would partly subsume the theory of externalities.) I will call this the Brainbent Theory.

Could this other half of economics be done by developing the "outsider" part of Coase’s theory of the firm?

(Sorry to get micro in a macro site, but you started it...)

That OTHER HALF would be ---> the Time Costs (in opportunity costs) of the Transactions Costs to participants, who are in market trades with private institutions, i.e. mainly business firms.

The fact that these ubiquitous opportunity costs are frequently non-monetized shouldn't reduce their real importance.

Example: The first step toward privatization of Medicare has resulted in seniors trying to figure-out which of 73 cards gives them the best deal, and then trying to stay current with that decision, in a changing market situation. Clearly, we reduce the amount of time they have available to figure out other things in their lives. In pure clock-time, it reduces available freedom.

In this case, a government program would save costs to the WHOLE system, by eliminating the continuous transactions costs to its participants. (And it might well out-perform prices in an oligopoly, once the cards shake out.) And productivity gains in this area would STILL continue, although not quite as fast, nor spread as quickly.

This could be formulated as the Brainbent General Trade-Off: (1) the population-distribution of liberty-time, versus (2) the velocity of productivity.

This trade-off would have a shifting boundary related to, among other things, the current distribution of income.

This theory would be PRIOR to the facts that in some situations, transactions costs, no matter how high, only lead to incomplete results (because of asymmetric information) or are overwhelming to contemplate (because the object is a public good): additional reasons for standard policies from the institution of government.

It was Coase himself who thought that his approach would finally rewrite all of economics. Maybe we start to do that by turning him upside-down.

Posted by: Lee A. on June 30, 2004 09:39 PM


Is this argument of relatively recent origin, or has it been around for a while? I've seen it a couple of times in the past year or so but I hadn't seen it previously. The version I saw argued that Pareto optimality (don't make someone better off unless it makes no one worse off), seemingly as broad and general and value-free as you can get, can actually be seen in this same light as a sop to the rich. I have to say that it smacks of sophistry to me.

Posted by: Half on June 30, 2004 09:56 PM


Thanks for welfare econ refresher. I think the point is important, and something the folks who think the market is a magic philosopher’s stone either don’t know or want to hide. But Prof B's statement is a hash on many levels and you have only commented on one of them.

First, GNP is a measure of production that results in final income to people (with some allowance for depreciation), not wealth. GNP leaves out so much economically relevant to measuring the stock of wealth in an economy (even from a strictly market economy perspective), I am not sure that a reasonable person would say that they will usually move in the same direction.

But lets assume Prof B meant income.

GNP also adopts a social production theory of value, and the value of social production is determined in the US by the market. So, the bottom line is that according to the GNP measure, if you expect to make $45,000 a year until you retire, at a 5% annual discount rate, your life at age 20 is worth exactly $809,145.71 (in today’s dollars). A minimum wage worker’s life is worth $185,204.46. If we want to get all gushy and soft, and grant these suckers a lower “social discount” rate of, say, 3% per year that they could never get on the market, then we have a princely $1,126,111.85 for the high pay person, and $257,754.49 for the minimum wage stiff.

I looked at his post. It was silly, in my opinion. Measuring floor space of the average European and US house to measure comparative welfare. That is similar to saying that slaves in the US were better off than in Africa, because they weighed more than African natives. I guess those geese flying all over the world really envy the inmates of goose liver factories. Why those lucky duckies even get to have nice man come over with funnel to help them eat. Wow!

Posted by: jml on June 30, 2004 10:59 PM


The Feldstein description of the Posner poistion is correct. Put another way, it's stupid to tinker with contract and tort law to try to help the poor (and sock it to the rich) when the leg can do the same with the stroke of a pen without most of the accompanying deadweight losses.

Posted by: alkali on June 30, 2004 11:04 PM


Actually, if we are hunting for the utility of wealth, the best proxy could be

U(i) = a(0)(W(i) - W(0)) + ... + a(n)(W(n) - W(i)).

In other words, wealth matters primarily as an indicator of status. Maximize that!

Posted by: a on June 30, 2004 11:36 PM


Reading this post made me curious why anyone would adopt this method of argument. Bainbridge assumes equivalence between private wealth and social welfare. Any claims he makes about social welfare are deductive and tautological.

That done with, a question for the economists..... Is Brad's argument contingent on the notion that all individuals share a generalizable utility curve, while Bainbridge's idea of utility implies the existence of multiple and incomparable utility curves?

Posted by: from a curious non-economist on June 30, 2004 11:52 PM


Two questions, Brad:

1) How do we know that individuals have similarly shaped utility functions? And if they don't, how would that affect your argument? In other words, could one make a case that your logic confuses cause and effect, that high-income earners earn high income because their individual welfare functions have a steeper shape, and so they have a higher marginal utility of wealth than low-income earners?

2) As a matter of practical policy, the question is not whether the free market's social welfare function is perfect, but whether it's better than the alternative. From your writing, I guess that your alternative would be a mixed economy with a welfare state that redistributes income through a political process.

With that in mind here's my second question: Assuming your individual utility function is a good approximation to reality, what is the political process's social welfare function? And is the utility distribution that maximizes it any less unequal than the free market's?

Posted by: Thomas Blankenhorn on July 1, 2004 12:56 AM


Two points:

(1) A wide range of social welfare functions are Pareto efficient. One could use a "Rawlsian" social welfare function, where

W = min(w1,w2,w3..........wn)

Now so long as utility is a montonic incresing transform of w, the shape of the function doesn't matter. This welfare function will NOT mean everyone has the same w, because incentives must be in place so that the smallest w is a large as possible. Once this happens, no other distribution can be make everyone better off .

Indeed, it is possible (although extremeley unlikely) that a strategy that maximizes total W also assures the best outcome for the worst off person.

(2) These social welfare functions assumes no interdependece of utility. This is obviously untrue: for instance one's utility is affected by his/her friends and family's utility. Some people worry about whether others are going to bed hungry. Others worry about whether their colleagues make more money than they do.

Moreover, recent literature on the economics of happiness shows that where people are relative to others in their societies matters to them--people seem to care per se about whether within their country they are better off than average.

Posted by: Richard Green on July 1, 2004 07:07 AM


John Emerson,

I think you hit a valid point: "some comparability of welfare is possible." Not always, but sometimes we can. The goal here is maximizing everyone's utility, where every utility is a separate function. Usually they can not be maximized at the same time. We have to make a "Multiple Criteria Decision".

To deal with such a problem we can either use:
1) Aggregation Functions (as shown above)
2) Tests for Synthetic Superiority with Accepted Incomparability.

The Aggregation Function is a mathematical tool, which allows for three conclusions: "==" , ">" or "<". The tests add a fourth conclusion, incomparability.

These tests are usually used to help one person, a decision maker, decide between alternatives where there are multiple criteria. The goal is to find the best alternative, not to measure it's utility. Applied to the case of utility maximization the utilities become the criteria. What you need is a decision maker, a "Just Judge". Obviously, that is the difficult part. Who would be the Judge?

The tests accept that sometimes you can not decide which alternative is better, for example whether to take or not $5 from a person earning $101,000 per annum and give it to the person earning $100,000 (extremely obvious example). OTOH, taking $5 from any of them and giving it to Malaria Prevention might seem a good idea to 99,9% of people/Judges.

In a decision situation whether to redistribute or not, as long as the test does not show Incomparability or Negative, it would be a signal in favour of redistribution.

An example of such a test:
B. Roy: (1968) : Classement et choix en pr'esence de points de vue multiples (la m'ethode ELECTRE), Revue Fran,caise d'Informatique et de Recherche Op'erationnelle 2, 8 pp. 57-75.

Posted by: MarcinGomulka on July 1, 2004 07:47 AM


Messrs "Blank" and "A":

The utility function being logarithmic has been assumed for the St. Petersburg Paradox as well; at first blush it seems eminently reasonable (and is also nicely tractable for any mathematics). I think it does start to get shaky when compared to actual human behavior - what were once luxuries become necessities, etc. A hat-tip to Mr. "A" for thinking of the jealousy factor as well: it surely must enter into any realistic assessment of utility.

I think Bainbridge's point may be a positivist one - it's easy to measure GNP or income or whatever, and difficult to quantify other aspects of utility. The offshoring debate touches upon this. One of the implicit assumptions seems to be that GNP increases are the main measure of the goodness of offshoring, but it ain't so.

Posted by: Bruce Cleaver on July 1, 2004 08:08 AM


I'm sure conservatives would take issue with the logarithmic utility function, but that's fine. What would they suggest? Linear? That means giving a millionaire $10000 has the same effect as giving a poor person $10000: try to defend that.

Posted by: RichK on July 1, 2004 08:37 AM


Why does utility enter as an argument in the wealth function?

Posted by: Michael Greinecker on July 1, 2004 09:35 AM


My real point is that I think that there is a sort of common-sense, folkish idea of "relative utility" which can be regarded as meaningful and valid even if not provable or rigorously quantifiable. Quantifications are fine if they work, but even without them it would be right to apply these folk judgements to policy-making.

The alternative is to allow the idea to be defaulted in that there are no comparisons between individual utility functions, so that it cannot be said that death from disease is worse than not being able to afford the best kind of cigar (stereotyep alert!).

The rejection of the utility function is **presented** as scientifically cautious and skeptical, but (especially because economics **is** a policy science) it amounts to begging the question and making an unwarranted positive assumption (**not** just a skeptical suspension of judgement) which really comes out of thin air.

Posted by: zizka / John Emerson on July 1, 2004 10:49 AM


"Why does utility enter as an argument in the wealth function?"

I suspect that this question comes from thinking of the functional dependence of wealth on utility as a *causal* dependence, which is not the right way to think about it.

Posted by: bza on July 1, 2004 01:33 PM


My uncle is a law professor. He thinks Posner's views are pretty bad law, generally.

Now I get to read Brad DeLong on it, which shows (at least, to my view), that it's pretty bad economics.

Law and economics: bad law AND bad economics!

BTW: Bainbridge's distortions aside, the Posner position also seems wacky to me. The idea that common law should orient toward maximizing wealth, and the legislature should worry about distribution, seems strange: I always thought that common law should be based on precedent, and if the legislature doesn't like it, for either wealth or distributional purposes, they can damn well pass a statute to modify the law, since statute trumps common law. I don't have anything against Kaldor-Hicks welfare analysis -- sure, it can be distorted and used for evil, but so can common law -- but if we want our laws to be based on it, then the legislature should implement laws of that type, not judges. Then again, Posner knows more about law than I do. I just... don't understand his view in the context of how, in the nuts-and-bolts sense, law comes to be. Any lawyers around here? Maybe I'll e-mail my uncle about it, but then, as I've mentioned, he's not a Posnerphile.

Posted by: Julian Elson on July 1, 2004 02:38 PM


It seems odd to me, too, that wealth-maximization should be a ruling principle of law. But it is, why shouldn't distribution also be a ruling principle. Both seem extra-judicial.

Posted by: zizka / John Emerson on July 1, 2004 02:58 PM


The real question is a bit simpler, it seems to me: can we reliaably measure the amount of happiness that particular activities give us and can we do it across cultural barriers? If so, then all the original arguments for using indirectly measured utility collapse. We could directly compare degrees of happiness accruing to us from particular activities and act accordingly.in terms of social decisions.
Lord Layard has recently asserted that the evidence is now available through measurement of brain states to allow us to do just that. I've yet to see a refutation of his position that makes sense, but this isn't my line of country. Still if we have a way to measure happiness on a ratio scale that works across cultures, then the thing to do is stop theorizing and start collecting data.
All this isn't meant as a slap at Brad. This entry is the kind of thing that makes reading his blog daily worthwhile.

Posted by: tracy lightcap on July 1, 2004 03:22 PM


I've been thinking recently about concepts related to the point that "a" brought up and Bruce Cleaver called the "jealousy factor", that some difference of wealth or income may be desirable "as an indicator of status". What I'd like to interject here is that that is merely the means to a means to an end. The intermediate means would be that the uneven distribution of wealth/income is what provides workers with the incentive to work hard and excel, in capitalist thinking, as far as I (also a non-economist, though I'm contemplating it as at least a minor) understand. And that, in turn, gives us the end of increasing (or even maximizing, we might hope) total production.
Could somebody tell me if I've got that more or less right? ;)

What I've been contemplating in particular is exactly what distribution curves, for both wealth and income, would give adequate incentive to want to move up the curve, while still distributing as equitably as possible. I've been looking at the Pareto distribution and its simplified form in the Pareto rule (not closely related to the aforementiond Pareto efficiency/optimality), and cursed the distribution when I realized it wouldn't be integrable to infinity. I'm sure that there have been papers written on the matter of what constitutes an ideal distribution curve, if not whole chapters of economics textbooks, but if anyone knows a site with a good summary and a couple of formulae, I'd much appreciate a link to it, either here or on the Wikipedia page listed below.

Posted by: John Owens on July 1, 2004 05:24 PM


Here's a half-hearted defense of Posner.

Suppose (under the usual assumptions) that wealth goes up by a little bit. Then there's a way to make everyone better off.

Suppose that wealth goes down by a little bit. Then there is no way to make everyone better off.

Of course, your welfare function could go down in the first case and up in the second!

Posted by: Hal Varian on July 1, 2004 07:10 PM


Brad has alluded to this argument before. I'm glad to see it worked out.

Posted by: Noumenon on July 1, 2004 11:06 PM


zizka wrote:

"My real point is that I think that there is a sort of common-sense, folkish idea of "relative utility" which can be regarded as meaningful and valid even if not provable or rigorously quantifiable. Quantifications are fine if they work, but even without them it would be right to apply these folk judgements to policy-making."

I think we could do a little better than that. I read recently a study that showed people's happiness went up with their income - up to about 40,000 a year, after which there was little further increase. This is consistent with the intuitive distinctions between necessities and luxories - even though those categories are not rigid, they are not non-existent.

Also, since we are nowadays abandoning the ole blank slate idea, we have a basis for saying that human nature is variable within parameters, not absolutely, and possibly a basis for asserting that one person's valuing a cigar more than another values life is outside the parameters, or at least outside the parameters of sanity, "insane" desires being ones we do not usually respect.

Maslow's hierarchy of needs might provide a good starting point for assigning values in a neutral way, but:

Tracy Lightcap wrote (in slight paraphrase):

"Lord Layard has recently asserted that the evidence is now available through measurement of brain states to allow us to reliably measure the amount of happiness that particular activities give us and we can do it across cultural barriers."

Want to maximize the population's serotonin levels? Make heroin available at cost. Toss in coke and ecstacy for variety. I'm not just being flip (though I'm not actually advocating this). The most effective way to create euphoric brain states is with drugs (well, sexual orgasm, but that's fleeting, and can also be enhanced with drugs), and if the recreational drug industry were brought aboveground, the technology would no doubt improve drastically. Whatever brain state Lord Layard considers optimal can be engineered towards. I think we need a more sophisticated definition of "happiness" than this, even at some cost to empiricism.

Regarding the jealously factor:

As I understand it, in the ancestral environment where human psychology was forged, those who acquired more goods than others did indeed get more status, but *by sharing the goods*. You used wealth to get status, but it was an actual exchange - you had to surrender the wealth to accrue the status. And if you did not, you might accrue club blows to the head. So you had status competition and redistribution. I think we today have various incompatible political instincts arising from the fact that values that were complementary in the ancestral environment come into conflict in more complex societies.

zizka also wrote:

"I think that a lot of the anti-ethical, value-neutralist tendency in academic life (and that's what Bainbridge's principle was) comes from the post-WWII revulsion at extremist moralizing philosophies like Marxism."

I think it's science envy, where science means natural science, and I think Marx was infected with it too. Science has achieved a lot by casting off religion and its values, and the other fields want to emulate that success and independence. Yes, Marx was quite the moralizer, but you can accept his analysis without embracing his values. Not many Marxists do, but I think that has to do with what sort of people are drawn to Marxism, not the logical consequences of the ideas. And no, I am not a Marxist.

Posted by: Martin Bento on July 2, 2004 01:08 AM


I can't believe Hal Varian posts on this board. I'm starstruck. . . .

Posted by: Bobby on July 2, 2004 01:15 AM


"I think that a lot of the anti-ethical, value-neutralist tendency in academic life (and that's what Bainbridge's principle was) comes from the post-WWII revulsion at extremist moralizing philosophies like Marxism."

There is no place for ethics in historical materialism. Marxism in its original form is definitely not about ethics.

"I suspect that this question comes from thinking of the functional dependence of wealth on utility as a *causal* dependence, which is not the right way to think about it."

Even then one would have to assume that different wealth levels would lead to different levels of utility.

Hal Varian has posted before on this board.

Posted by: Michael Greinecker on July 2, 2004 01:38 AM


Yeah, Bobby: more influential people than you might imagine read this blog.

Posted by: Richard Cheney on July 2, 2004 07:56 AM


I would like to argue that Posner as interpreted by Feldstein and Musgrave is still clearly dead wrong.
He could have argued that "common law judges" should get us to (and keep us on) the feasible welfare frontier, while legislators decide how to distribute. He must not argue that others (including judges) should maximize wealth and then legislators should distribute it. He certainly agrees that redistribution is a leaky bucket due to incentive effects. This means that less total wealth distributed in a way acceptable to legislators means more total wealth in the end.
The division of labor does not make sense because there is no total wealth neutral way to redistribute. Pretending that there is (and Posner sure isn't unusual in doing so) while arguing against redistribution by any feasible mechanism is intellectual fraud.

Posted by: Robert Waldmann on July 2, 2004 03:56 PM


According to David Fridmans book "Laws Order", Posner helds the view that common law is already mostly wealth maximizing.

Posted by: Michael Greinecker on July 3, 2004 07:41 AM


Gee I guess I missed all the fun these past few days. Though, I'm scratching my head to figure out what all the fuss is about. I think we need some positivists.

If understand correctly what is being discussed here, i.e. Suppose there is somehow, someway to assign a notion of wealth W_i per individual, then maximizing \sum_i W_i is a social welfare policy that avoids ''interpersonal comparisons of utlities''? Well how do we measure this notion of wealth -- certainly not in monetary terms, and some "real output" production-function type model seems to assign soviet-style measures of wealth.

Doesn't this bring is right back to the same operational problem that one would face for utilities? I'm confused.

BTW Delong I'm not sure how you produced those formulas (if it's LaTeX I sure don't see it), but they look awful.

Posted by: CSTAR on July 3, 2004 11:33 PM


Great site fatty lose weight with reductil and reductil uk

Posted by: reductil on July 6, 2004 02:39 PM


Great site fatty lose weight with reductil and reductil uk

Posted by: reductil uk on July 6, 2004 11:02 PM


Great site fatty lose weight with reductil and reductil uk

Posted by: reductil uk on July 8, 2004 03:22 PM


Get it up mate, it's fun!

Posted by: Viagra on July 8, 2004 09:12 PM


This will get yours up again, dude!

Posted by: Viagra on July 12, 2004 03:09 AM


It gets yours up to the top dude! The girl will enjoy it!

Posted by: cialis uk on July 13, 2004 07:20 AM


Muppets love Viagra!

Posted by: Viagra on July 14, 2004 04:29 AM


I don't really think your thoughts are right. Maybe you need a loan?

Posted by: payday loans on July 15, 2004 01:12 PM


I don't really think your thoughts are right. Maybe you need a loan?

Posted by: payday loans on July 15, 2004 08:02 PM


Hire a car to feel higher!

Posted by: Car hire on July 20, 2004 05:27 PM


Hire a car to feel higher!

Posted by: Car hire UK on July 21, 2004 12:06 AM


Great site fatty lose weight with reductil and reductil uk

Posted by: reductil on July 23, 2004 02:05 PM


Post a comment