More on inequality …

by Daniel on December 19, 2003

Just “feeding the baby” with a couple of links really …

Stuff from Maxspeak, Paul Krugman and Calpundit relevant to our own discussion of “Equality of outcome versus equality of opportunity”. Read them all. (If you want to that is, I mean it’s not like I’m ordering you to read any of them or even suggesting that you’ll be materially less well-informed if you don’t. I’m just sort of suggesting that they might be a little bit more interesting than what’s in the newspaper today)

Personally, I’ve always had a hard time taking this debate seriously. Specifically, I’ve never received (not for want of asking) a satisfactory answer from anyone who talks about “equality of opportunity” to the following two questions (also inspired by my time at business school, which I am coming to believe may have been less wasted than it seemed to be at the time)

1. What’s the point of doing anything if you’re not going to check whether it worked or not?
2. How do you find out whether a course of action worked or not, other than by the results?



John Q 12.19.03 at 9:34 am

Thanks for the links, D^2.

As the Krugman link makes clear though, I don’t think that your implied claim that equality of opportunity is untestable in the absence of equality of outcomes can be sustained.

If outcomes were uncorrelated with all the obvious measures of starting point (parental class, race, gender etc) it would seem reasonable to conclude that opportunities were substantively equal.

Of course, this isn’t the case, and the evidence shows that the more unequal the outcomes in one generation, the more unequal the opportunities for the next. Again, Krugman has this nailed.


dsquared 12.19.03 at 10:58 am

Hmmmm yes I suppose so. Now that I come to think of it, I originally came up with that joke in the context of a discussion of racial equality rather than social. ah well …


Chris Bertram 12.19.03 at 11:17 am

I think you might want to limit the scope of 1. a bit. I imagine that you or some of the people who came out with this dictum at business school are probably going to want to bequeath some of your possessions one day.


dsquared 12.19.03 at 11:32 am

ach, what did posterity ever do for me?


Anno-nymous 12.19.03 at 11:47 am

Regarding the opportunity/outcome debate, I think it’s worth mentioning that even if opportunity, in the real world, necessarily implies outcome — I think many of the outcome folk argue this point — opportunity is still more defensible as a philosophical proposition. In my opinion. And almost certainly in the opinion of any political conservative.

And hey, if you can empirically convince me that my belief –and conservatives’ — in equality of opportunity morally impells me to fight for equality of outcome, so much the better.


Andrew Boucher 12.19.03 at 12:48 pm

Is chess a fair game? Does one look at results to decide whether it is or not? No, it’s the rule book which is important.

Sure some people play chess better than others. Also, suppose there are two equally talented people, but one has a chess-playing parent and the other not. Then I’d wager the first will probably play better – he gets more practice for one thing. Still, chess is a fair game

Auditioning for orchestras by having a system where people play behind curtains is fair. Even if the results are skewed – as I imagine they are – to people who grew up in a family with musical background..

Fairness of opportunity can be checked by checking the rules. If the rules are fair, then opportunity is equal.

This is not to say that I think there is (today in the US) perfect or even near-perfect equality of opportunity. But I certainly wouldn’t want to live in a country where equality of results is the metric by which one measures whether the system is working or not. Is that what you are advocating??!


Tom T. 12.19.03 at 1:10 pm

It may be that you have not gotten a satisfactory answer to your question because the people to whom you have posed it do not share the assumption built into it. Of course one should check the results of one’s actions. Your question, though, assumes that equality of opportunity only “works” if the result that it produces is equality (i.e., equality of result). That’s question-begging, and it is not unreasonable to think that one who argues for the equality-of-opportunity side is not going to buy into it.


dsquared 12.19.03 at 1:27 pm

If the rules are fair, then opportunity is equal.

Do try to keep up., Andrew, this is the exact subject we’ve been discussing for days. Not true; opportunity can also be unequal if initial endowments are unequal.

And Tom:

Your question, though, assumes

No it doesn’t. It’s simply an invitation to anyone who craps on about “quota systems”, “reverse racism” or mentions that Krt Vonnegut story, to put some substance behind their lip-service of “equality of opportunity” and suggest how they’d carry out a policy to equalise opportunities. JQ pointed out an alternative (whether the status of successive generations of the same family was uncorrelated), which is good as far as it goes, but not general in its application because you can’t apply it to gender or racial inequality. Annonymous above understood the point of the issue; the theoretical debate on oportunity versus result is separate from the practical debate.


dsquared 12.19.03 at 1:39 pm

Just on a theoretical point that occurred to me over lunch; I don’t think it’s possible for outcomes to be “uncorrelated” to race and gender unless the mean outcomes for different racial groups are equal to the mean outcome for the population as a whole, which is what I was trying to get at.


Mikhel 12.19.03 at 1:58 pm


In somewhat keeping with the analogy, we’re examining how many pieces people get at the starting point. Those of us on the right of the issue don’t fare well in a chess analogy (that I can think of at the moment) so I’ll leave it alone.

Thanks for the links, Mssr. Squared.


Roger 12.19.03 at 2:06 pm

If you want equal outcomes, just kill all the rich and the poor and the remaining people will starve to death in perfect equality. (In practice, we don’t kill them, but we do punish them.) What we should seek is optimal outcomes, not equal outcomes. If a system makes my neighbor 100 times richer than me, but makes me better off than a competing system, then I should prefer the first system. Seeking equal outcomes is pure envy, something more base than greed. Freedom is the optimal system but it has a hard time competing against special interests in the political market.

[DD as moderator: email address removed following request. Be careful out there people]


Mikhel 12.19.03 at 2:45 pm

I was reminded of an old article in The Observer after reading Krugman’s piece. It’s called, “From the Log Cabin to the Whitehouse”, and appeared in 2002. is hosting it now; the link is below. I think many of the people who visit this blog are from the UK, and I know some of the bloggers read The Observer. Forgive me if it’s been posted before.


Matt Weiner 12.19.03 at 3:35 pm

If a system makes my neighbor 100 times richer than me, but makes me better off than a competing system, then I should prefer the first system.

We’re at least halfway to Rawls’ Difference Principle now–what yields the best result for everyone, or (thinking from the Original Position) for the worst off? I’d guess, and IIRC Rawls would agree with me, that this question leads to economic policies considerably to the left of what exists in the U.S. today.


amelia 12.19.03 at 3:43 pm

Disclaimer: I haven’t followed the debate before this thread. Too much day job.

Maybe a more interesting question is whether and how we can make broadly applicable rules that actually lead to substantive autonomy for all people (a better measure than “equality,” I think).

I would also disagree, on substantive autonomy grounds, with roger’s contention that “freedom” (unfettered market capitalism?) is “the optimal system.” How is it best? What are the empirical, as opposed to the theoretical, results for the worst-off?

Look to, for example The Real Worlds of Welfare Capitalism.


Jon Mandle 12.19.03 at 3:50 pm

Rawls occupies a position in this debate that is more interesting and subtle than is often appreciated (not directed at you, Matt). His second principle of justice has two parts: fair equality of opportunity and the difference principle. The fair equality of opportunity principle is obviously concerned with … well … opportunity. It says roughly that individuals with similar capacities and motivation should have equal chances of occupying the various social positions generated by the basic institutional structure of society. But this leaves open the question: what are those social positions that are to be filled fairly? The difference principle helps to answer that question. It says that the institutional structure in which the worst position is better than the worst position in any other structure is the best one. So systems are assessed in terms of their likely structural outcomes. But once a just structure is in place, we do not enforce any particular outcome. Individuals will make their way to various social positions based on their hard work, luck, the supply of and demand for their talents and skills, etc., against a fair set of background rules and fair equality of opportunity.


ws 12.19.03 at 4:22 pm

To answer question 2, let’s look at satisfaction with process vs. outcomes. If people perceive that they have had a fair deal with the process, they will be less likely to be dissatisfied with adverse outcomes. So making sure that the process is fair is more important than bothering about the results of actions.


Andrew Boucher 12.19.03 at 4:24 pm

“If the rules are fair, then opportunity is equal.”
Do try to keep up., Andrew, this is the exact subject we’ve been discussing for days. Not true; opportunity can also be unequal if initial endowments are unequal.

Well then you’ve just changed what people usually mean by “equal opportunity”. Most people – including me – would be satisfied if the rules are fair and mean “equal opportunity” in this way. If you’re meaning it in some more literal sense, then go ahead; but then any conclusions you make would apparently be about something different than what most people mean by “equal opportunity”.

In somewhat keeping with the analogy, we’re examining how many pieces people get at the starting point.

Maybe but the analogy doesn’t work. The rules of chess state how many pieces people get at the starting point – they both have the same 16. If any player gets a different amount, then he is following different rules.


Dick Thompson 12.19.03 at 4:28 pm

Dare I mention IQ? (Covers head and waits for storm).


dsquared 12.19.03 at 4:29 pm

Most people – including me – would be satisfied if the rules are fair and mean “equal opportunity” in this way.

So if I entered an arse-kicking contest against a one-legged man, most people would say that he had an “equal opportunity”? Do me a favour.


dsquared 12.19.03 at 4:31 pm

Dick: Well of course you may mention it, but you’re probably going to have to give us a bit more to go on if you want to bring it into the discussion.


Katherine 12.19.03 at 4:53 pm

I think as far as policies go the debate in the U.S. is not between people who believe we should have equality of opportunity versus equality of coutcome, but between:
1. People who say they want “equality of opportunity” but think this is satisfied as long as the government treats everyone the same. (Schools are exempted from this with the “local control” excuse. The quality of legal defense for poor people? Well, it’s not like the government pays for Johnny Cochran) In that case, the results don’t matter.
2. People who want equality of opportunity–or to get much closer to equality of opportunity–on the ground, in fact. There, it does become intertwined with equality of result.


Matthew 12.19.03 at 4:55 pm

On equality of opportunity I think the story that says it all is,

‘It’s each animal for himself’, said the elephant as it danced among the mice.


Matt Weiner 12.19.03 at 5:38 pm

Andrew–“Initial endowments” by me includes money as well as natural endowment. I don’t think a situation in which, say, army commissions are for sale would count as “equal opportunity” for the poor, but it would treat everyone the same. The law in its majesty forbids the rich as well as the poor to sleep under bridges, ‘n’at.

(Irrelevant digression: Chess is sometimes played with odds–one person starts without a rook or something–and ironically in this context White has an advantage over Black by moving first.)

Jon M.–Thanks, feel free to correct me on Rawls if it seems necessary; I don’t claim him even as an AOC. :-)


Andrew Boucher 12.19.03 at 6:05 pm

So if I entered an arse-kicking contest against a one-legged man, most people would say that he had an “equal opportunity”? Do me a favour.

And so two people auditioning for an orchestra by playing behind a curtain so that the only thing which matters is their music – that’s not fair just because one was raised by musically-inclined parents? Do me a favour!


jimbo 12.19.03 at 6:22 pm

Without mentioning the Evil Book Which Must Always Be Denounced by name, I’ve always been interested the fact that liberals have not used it to argue for compensatory measures.

After all, one of it’s central themes is that we are manifestly not created with equal endowments, and that some people through no fault of their own are less able to take advantage of any opportunities that they are given. So why not take that as an arguement for cushioning the extremes of a capitalist market?


dsquared 12.19.03 at 6:26 pm

Andrew, you keep equivocating between “equal opportunity” and “Not fair” and I really must request that if we’re going to get anywhere in this discussion, you’ve got to stop. They don’t mean the same thing and one, but not the other, is a value judgement. Replace the words “not fair” in your example with “unequal opportunity” and it doesn’t work.


Decnavda 12.19.03 at 6:36 pm

I think you inappropriately conflating practical tests of equality of opportunity with a need for equality of outcome.

Rawls used gambling as an example of system with perfect proceedural justice: As long as the rules and process are fair, the outcome is just. But how do you know if a particular outcome is fair, if the dice or roulette wheel are baised? You test them, looking at the results of many throws and spins. But even if the system is perfectly fair, the resulting just outcomes can be extremely unequal.

And yes, we can apply this, to say, race. There is good scientific evience to indicate that skin color has no corolation to IQ or talent. So if the results show blacks disproportionately failing, we can assume the system is unfair. But does this mean that we should try to equalize outcomes? Maybe as a temporary remedial measure, like giving losings back to a gambler when the dice turned out to be biased, but I do not see why this means that equality of outcome should become any sort of goal.


Sebastian Holsclaw 12.19.03 at 7:51 pm

Until you define how you want to check things, your question #1 is useless. If you are checking things by saying, are the outcomes equal you are presupposing that equal opportunities will lead to equal outcomes, which completely ignores individual choices about work, study, etc.

Regarding the difference principle, I’m not at all convinced that it is a good guide.

Take society I with positions A, B, C and D where A=500 goodness units (GUs), B=499 GUs, C=498 GUs, and D= 250 GUs. Assume a 1/4 distribution in each from a population of 10,000.

Now take society II where A=10,000 GUs, B=5,000 GUs, C=4,000 GUs and D=249 GUs. Population distribution in society II is A=2500, B=3500, C=3999 and D=1. The difference principle suggests that Society I is better even though Society II is better by two full orders of magnitude for 9,999 people out of 10,000 and is only slightly worse for 1 person out of 10,000. It isn’t at all unreasonable to assume that even with a ‘Veil of Ignorance’ people might choose society II because their chances of being in D in society II is 1 in 10,000 and they are only marinally worse off than being in group D in society I which offers a 1 in 4 chance of being in group D.

Note also that society II has a much less even distribution of GUs even though nearly every member of the society is 100 time better off than in society I. But that is a slightly different discussion.

I also note that we are pretending that society ‘distributes’ everything to people as if people were unable to ‘earn’.


Andrew Boucher 12.19.03 at 8:08 pm

I’ll make this last comment and stop here. I’d say your arse-kicking contest is a bad example because it depends on the pity factor. Take a reasonably close example: the NBA. Is the NBA an example of equal opportunity? Even if the results of who gets to play in the NBA is loaded with unequal results? (skewed to blacks, for one thing). If you say it’s not an example, ok, but then your meaning of “equal opportunity” is not so pertinent to its common usage, because when most people think of “equal opportunity” the NBA is a prime example – the only thing that matters is how well you can play basketball.

But could be missing the point of what you’re saying completely. If so, apologize in advance.


epist 12.19.03 at 8:22 pm


As to the distribution difficulties of the Difference Principle, I agree that it needs some tweaking if it is going to satisfy the bulk of our intuitions concerning choices between societies from behind the veil. The one bit of statistical data I’ve seen on this suggests that people are much more likely to favour a ‘minimum threshold guarantee’ than the unadulterated difference principle. Note that, on your example, poor D falls well below any minimum threshold (how do you see a 25,000% reduction in goods as ‘just a little worse’, anyhow?) in the second scenario, with his measly 1 unit. Of course, you can rig the example to out the threshold intuition by merely rasing the bottom bar. But the threshold society is no more libertarian than the difference principle ociety, as both are premised on the mutual agreement of those in the original position, and not on (for example) the prior existence of things like property, rights, or desert.

And as to the complaint that liberals treat all ‘goodness’ things as the property of the government to be distributed, I sympathize with your (and Nozick’s) complaint here, but I think it’s based on a misunderstanding of what the claim is. The liberal claim isn’t that all good things are the government’s to distribute, but rather that, in a society, people have and enjoy their good things only if the society permits. For the argument behind this (and a devilish use of the claim of actual equality) I point to Hobbes, who argues that people are so nearly equal in all abilities, and so very violent and vengeful, that they can put an end to any enjoyment by any man they choose. This seems to me about right. So, as a practical matter, the society, by deciding what it will and will not allow, ‘distributes’ all good things. Not that it (the gov, or whoever) makes or delivers these things, but that it regulates their possession and enjoyment.


carlos 12.19.03 at 9:35 pm

In fact, the NBA has an interesting example on this subject; the draft. The draft is the most clearly redistributive measure you can think of and still only a small number of teams have won titles. Imagine if the draft were perfectly random, Boston and LA would have doubled their championships easily. This is a situation were “fair” rules create a clearly skewed result and only redistribution can force a reasonably fair result. Equality of opportunity in the sense of “fair rules” doesn’t mean much when the results tend to be self reinforcing ie: the son of a upper middle class guy has a bigger chance of getting richer than a poor guy’s son of becoming middle class, so in the long run the distance and rigidity of social classes seems to grow (all things being equal)


novalis 12.19.03 at 10:39 pm

decnavda, apropos your comment about randomness, a Dilbert comic (this may only be funny to people who know something about math).


Zizka 12.20.03 at 12:04 am

1. Unsuccessful people who believe that life has been unfair to them want life to be fair. Those who have succeeded under conditions unfair to them might feel especially proud of their own success and be hard-hearted toward the unseccessful, or they might believe in making things fair. People who succeed under conditions biassed in their favor usually take their success for granted.

But almost everyone wants their own kids to have the maximum advantage, without regard for fairness. Almost no one says “If my son can’t make it fairly and on his own, I don’t want him to make it”. The test for real committment to equality of opportunity would be whether or not someone would make that statement.

2. A defining principle, perhaps THE defining principle, for American conservatives is the conviction that some people deserve to lose and to suffer. The losers might be sinners (Christian version), lazy, untalented, or unproductive (law of the jungle freemarket version), or suckers (Mafia / Halliburton version). The correct attitude toward losers is not sympathy, empathy, or pity, but schadenfreude. Serves them right — better them than me.

3. The losers in a game of musical chairs probably are the ones who most deserve to lose, but maybe it would be better to have enough chairs for everyone.


Matt Weiner 12.20.03 at 12:23 am

Agree with your comments about the minimum threshold, but you’ve misread Sebastian’s case slightly; in society D has 249 goodness units, and there’s only one person in D. (As a Rawlsian textual note, IIRC in the original position you don’t know how many people are in each group.)

You’re dead on about what it means for society to distribute positions. As Jon Mandle points out, the idea is that you set up a societal arrangement and let it rip, with everyone having equal access to positions given their character and motivations. You want the arrangement that’ll get the best possible outcome–that’s most unlikely to be done by setting targets and forcibly redistributing until you get those targets.

One thing to keep in mind is that these are Goodness Units we’re talking about, not dollar amounts. Giving A, B, and C lots more money and D the same amount will reduce D’s Goodness, because A, B, and C will bid up the price of the things D wants to buy. (Katherine pointed this out on the very first Sen thread; it’s also why, as Brad DeLong pointed out, perfectly functioning markets maximize a weighted sum of utilities, with more weight being given to richer people.

I’d be interested to find out if any economically literate people know whether unequally distributed productivity gains can theoretically diminish the utility of the worst-off in a perfectly functioning market–that is, if it can be that in society II D’s wage can purchase as much productivity as in society I, but D winds up worse off because A, B, and C can purchase even more productivity than in society I.


Kilroy Was Herfe 12.20.03 at 12:38 am

Sebastian —

It’s been awhile since I’ve read my Rawls, but I seem to recall that the Difference Principle is justified from the maximin principle in economics. The maximin principle only obtains in particular situations. Primarly, the agent has no knowledge of the likelihood of possible outcomes.

In your example, the agent has exact knowledge of the possiblities of outcomes (assuming random assignment).

Ergo, the justification for the use of the Difference Principle would not apply.

A better example would be the following:

We have two societies, the Nimrods and the Dumasses. In Nimrod society there are only three positions, King, Merchant, and Food. Kings have the greatest amount of social goods (on the order of 10000 GUs of social goods). Merchants have some social goods (100 GU of social goods). Food has no social goods. In fact, Food is tortured, beaten, and when finally rendered useless for pleasure by Kings or Merchants, eaten. (-1000 GU of social goods)

In Dumass society, we only have two classes of people, Citizen and Taxpayer. All citizens have 50 GUs of social goods. While Taxpayers have 5 GUs of social goods.

Now, obviously, it’s better to be a King or a Merchant than a Citizen or a Taxpayer.
Unfortunately, there is one very important piece of informaiton missing from your decision. You don’t know how many Kings, Merchants, or Food there is in Nimrod society. And you don’t know how many Citizens or Taxpayers there are in Dumass society.

Thus, you don’t know the probabilities of being assigned to a particular class.

Given this constraint (which is a constraint that Rawls places in his Original Position), if forced to choose between the two societies, the rational decision would be to maximize your minimum and choose the Dumasses over the Nimrods.

Hope this helps,
Kilroy Was Here


Kilroy Was Here 12.20.03 at 12:54 am

Most people – including me – would be satisfied if the rules are fair and mean “equal opportunity” in this way.

Ah, Andrew, but here’s the kicker. How do we know if the rules are fair? In the US, we assumed that the Civil Rights legislations in the mid-60s finally made the rules were fair for minorities. But how do we know? It depends on what we want to accomplish.

I personally look the great Martin Luther King for the goals of this legislation (and all legislation aimed at “leveling the playing field.”)

“I have a dream that my four children will one day live in a nation where they will not be judged by the color of their skin but by the content of their character.”

Civil rights legislation was an attempt to do this. So far, it hasn’t worked.

We still live in a society where minorities are more likely to do more time than whites for the same crime.

We still live in a society where minorities are less likely to get a job than whites with the same abilities.

We still live in a society where the homes of minorities are worth less than whites with the same abilities.

We still live in a society where the children of minorities are more likely to die in childbirth than the children of whites.

And on and on…

So, Andrew, if the ‘rules are fair’ for minorities, and in particular, blacks. How do you account for the difference in results?

Kilroy Was Here


Sebastian Holsclaw 12.20.03 at 12:56 am

Kilroy, if you are right, the whole concept is useless to me. A thought experiment about ordering societies that can’t differentiate between a society where 99.99% of people are well off and one where 0.01% of people are well off seems like a waste of time. But what do I know. I think that it is a waste of time to deal with economic theories which can’t deal with the idea that a stratified but generally rich society might be better than an equal but generally poor society.


Kilroy Was Here 12.20.03 at 1:17 am

Sebastian —

You bring up the point exactly. Rawls had to order societies and not know the exact probablities.

In his Theory of Justice, Rawls was concerned about justice in the ideal, rather than justice in the application. Rawls was comparing (and ranking) three conceptions of justice–utlitarianism, intuitionists, and Rawls notion of ‘Democratic Equality’.

If you had to order societies that used these three forms of justice, how would you do it? Rawls argued that you would do it through a thought experiment. What society would disinterested rational agents agree to in a position where they would only know the basic principles of each society?

Why did Rawls not worry about the specifics such as you brought up? Because his goal was not to determine the best society but to determine the best principles for justice.

The Theory of Justice is really a very wonderful book. It’s worth a read at least once a decade.



piraisoli 12.20.03 at 3:01 am

Sebastian Holsclaw: “..a stratified but generally rich society might be better than an equal but generally poor society.”

Upto a point, this is true, but not always. I agree that a stratified society in which no one starves is better than a more or less equal socirty in which a lot of people starve. But once a certain measure of this kind of good is reached, it is not clear that equality wouldn’t have advantages over stratification. Partly this is a matter of deciding what the dimensions of the good life are. I think in the West we have reached a technological capability for feeding, clothing and sheltering all members of our society in such a way that physical privation need no longer be an issue. Given this level of technology and social organization, I don’t think it is at all obvious that we should prefer society A in which the *average* wealth is, say, 1 unit, nobody starves or is cold, but most of the wealth goes to 90% of the people versus a society B in which the average wealth is 1/2 unit, but that is much more uniformly distributed, and also one in which nobody starves, etc. and in fact in both societies culture, science, etc., as we know flourish.



cure 12.20.03 at 4:17 am

The bigger problem with Rawls, as I believe someone mentioned earlier, is that the groups in a just society are often read as a distribution rather than as a result.

Imagine two societies: One has average wealth 1, great stratification, but there is “equality of opportunity” in the sense that it’s possible to be successful depending on one’s motivations and actions rather than on one’s birth. Chess is an apt example of this – some people are naturally gifted to play chess, some have chess-playing parents which helps them play better, but nonetheless it’s quite possible for anyone of any background to learn the game and play at a high level if they so desire. In this society, most, though not all, members of the “poor” class are poor because of their own desires, work ethic and actions.

Admittedly, some people are just unlucky. But there is nonetheless enough “equality of opportunity” for the vast majority of the poor class to pull themselves up, even marginally.

This is, I believe, an apt description of the goal of most United States policy, and a societal description that I think would be overwhelmingly favored by Americans should they have to choose from behind the veil.

Society 2 has average wealth 1/2, little stratification and very little in the way of poverty. One’s motivations and work ethic are still correlated with one’s success, but it is possible to live a decent life without such traits. There is a very real limit as to how successful one should be, and this society doesn’t really value those who work tremendously hard in an effort to exceed the mean. There is a similar amount of “equality of opportunity” as defined above, but there is less opportunity overall.

I still think one is a better society. I do think that hard work and correct choices should be rewarded by a market, sometimes heavily, even if the price to pay is that sometimes luck or, on occasion, birth, play into that success. I do think a just society should punish, at least on some level, potentially productive members who choose not to be productive. If one is able to work but chooses not to, I don’t think society is obligated to provide it with housing, food and clothing, and I can’t think of many non-Rawslian theories of justice that would disagree.

Look, equality of opportunity quite clearly means something far different from equality of outcomes. Approaching Equal opportunity means that approaching a society where productive processes are rewarded equally no matter who the producer is. Looking at outcomes is a terrible way of measuring opportunity since outcomes can never show desire, work ethic, intelligent choices nor sloth, negative affectivity and lack of desire.

As an aside, what’s with the “Goodness Units”? I know the problems with assigning real numbers to Utilities, but utils are still a mighty useful comparative measure in this type of ideological discussion. We’re all converting “GU” into utils in our head anyway, aren’t we?


anon 12.20.03 at 9:33 am

As an aside, what’s with the “Goodness Units”?

Amen. Found on Gene Expression:

To me the greatest problem with utility theory is that it generally ignores that after a base level of consumption is acquired, most utility is from relative status. This is why indices of ‘happiness’ from surveys show that we are no happier today than we were 50 years ago, in spite of massive increases in material wealth, leisure time, etc. Relative status makes sense to evolutionary biologists (get those genes in the next generation), but it’s anathema to economists because it makes their assumption of utility maximizing agents go from amorally selfish to immorally selfish (at least before the assumption was people were indifferent to others but their desire to be rich would indirectly lead them to a happy endgame, with relative status as a goal strife is eternal).

(at )


Roger 12.20.03 at 3:32 pm

Healing Our World by Dr. Mary Ruart explains well why an unfettered free market is an optimal system for everyone.


Roger 12.20.03 at 3:35 pm

My mistake, should be spelled Dr. Mary J. Ruwart,


Zizka 12.20.03 at 4:23 pm

A place where relative vs. absolute utility is absolutely clear is in an American highschool. Not having your own car is endurable unless a bunch of other kids have cars.

TV projects relative deprivation to the whole world anyway, but if it’s right there physically it’s worse.

Too mediocre a school might lead to limited aspirations, but in a very stratified high school many kids just get angry or demoralized and just give up. The kids at the top have too many advantages in all competitions.

The idealized American high school is a bunch of poor kids who are aspiring and working hard. How common these are I don’t know, but I think it works best without a favored elite group at the top.


Sebastian Holsclaw 12.20.03 at 6:54 pm

Unless I read it somewhere else, I think I made up the term Goodness Units for my hypothetical. I didn’t want to suggest that money was the only metric worth analyzing. I suspect there is some generally accepted term in the field, but I have no idea what it is so I made one up. (Thats what I get for mostly reading older philosophers).

I think most Americans would rather live in a society where there was a high (though of course not perfect) correlation between effort and success which allows for stratification. On this board the argument seems to be:
A: We don’t have that.
B: Therefore we should give up and try to equalize a lot of wealth because that seems fairer to try to erase both unearned and earned differences instead of trying to reduce the influence of unearned differences.

I obviously don’t believe in the ‘therefore’. I think it is a dodge to avoid dealing with the fact that quite a few people don’t share a moral vision where avoiding inequality of outcome is worth making massive changes to society.


Matt Weiner 12.20.03 at 8:08 pm

Goodness Units seemed perfectly clear to me–utils is what was used for it upthread. Of course one problem is that there is no such thing.

“a high (though of course not perfect) correlation between effort and success which allows for stratification.”

Now, does this mean that we ought to ensure failure for those who don’t try hard enough? What level of failure, exactly? Keep in mind that we’re not saying everyone must be made equal–reducing inequality doesn’t mean that.

Anyway, we’ve been discussing decreasing intergenerational social mobility in the U.S. That means that the stratification doesn’t reflect effort as much as it reflects mommy’s and daddy’s wealth. (At least, we’re arguing that you can’t tell that rich people’s kids try harder.) So the idea behind reducing inequality is not to give up on the goal of equal opportunity, it’s to promote equal opportunity by reducing the importance of unearned differences.

Of course, there’s a tradeoff–one generation’s earned difference is the next generation’s unearned difference. So this goal can’t be obtained without doing a bit of taxation of the rich. If letting this generation’s richest keep all their gains means abandoning the potential of the next generation’s poor, then I’m against it. Because that’s not a stratification that comes from the correlation of effort and result.


Matt Weiner 12.20.03 at 8:11 pm

Hm, “there is no such thing” came off as nasty, and I didn’t mean it to. What I mean is, there’s no perfect interpersonal and quantitative comparison of utility; it’s necessary to talk this way sometimes to make any sense (there was nothing wrong with Sebastian’s original example), but it’s an oversimplification.


Sebastian Holsclaw 12.20.03 at 10:50 pm

Sure it is an oversimplification, but it is an oversimplification shared by all theories on social organization that I’m aware of. So from the point of view of discriminating between theories it is an oversimplification that doesn’t matter.

I also don’t see why I should accept that because we are in a culture that allows even the poorest to have good food, good shelter, and typically some electronic devices that therefore we must suddenly become worried about equalizing outcomes above that level. It seems to me that leveling outcomes is much more important when the poor end up starving, or freezing etc. Then the stratification is causing physical misery so an appeal to alleviating that misery is likely to be broadly successful. But the argument can fall apart if you aren’t careful when even the lower income percentiles can own TVs and DVDs.

We can have a different argument about whether or not the US is actually at that point, but as a theoretical matter it isn’t obvious to me that redistributive arguments become stronger when a society gets past subsistance. In fact, I suspect the arguments become weaker to most people in such an instance.


Matt Weiner 12.20.03 at 11:59 pm

Sorry, I wasn’t clear the second time either. I agree entirely with your first paragraph, and never meant to be criticizing you at all on Goodness Units or utils.

As for the second–it’s not obvious that the poorest groups in the US are at acceptable levels–I refer you to the post on Sen on African-American life expectancy, which started all the flame wars. Our stratification seems as though it may cause misery even if the poorest aren’t starving or freezing.–Perhaps this is what you mean by the “different argument.”

But still, I don’t think you’ve addressed the argument that one generation’s fair inequality of result is the next generation’s unfair inequality of opportunity. I can’t speak for most people, but I’m quite uncomfortable with the idea that some people begin life with great shots at good jobs and leadership positions, while others, to attain those jobs and leadership, have to work their keisters off and catch some lucky breaks on the way. And that’s what limited intergenerational social mobility could mean.

One way to solve that would be to ensure that everyone is in the same place to begin with–that would be flat-out equality of result. Another would be to make sure that the processes by which people wind up in different social stations are more amenable to the worst-off. And that would involve a lot of government activism too, which would have a redistributive effect. (Affirmative action is a band-aid compared to what’s needed.)

I don’t think it’s feasible to ensure that everyone has an absolutely equal opportunity, and if it were feasible the particular policies probably wouldn’t be desirable. But we could, and should, do better. This isn’t redistribution for redistribution’s sake, but for the sake of, um, the children.


David Brake 12.22.03 at 4:50 pm

Getting back to the original articles for a sec – The Business Week article Krugman refers to is no longer available for free online however another blogger has provided the text on his own site: I have emailed David W. Wright to try to locate his study referred to by BW which “found that sons from the bottom three-quarters of the socioeconomic scale were less likely to move up in the 1990s than in the 1960s.”

The main point of trying to get more equality of outcome IMHO is that happiness (above a survival threshold) has been shown to relate much more to relative prosperity than absolute prosperity.

Also see several postings on the subject of the science of happiness from my weblog, notably:

Comments on this entry are closed.