I think I have mentioned that I shall be giving a talk to
the UNC Chapel Hill Philosophy Department in five weeks with the title “A
Game-Theoretic Analysis and Critique of John Rawls’ A Theory of Justice.” The
talk is pretty well prepared in my head [I wrote a book about the subject forty-two years ago, after all], but one small matter that is new in the talk has been
bugging me, and I thought I would talk it out here. What is at issue actually is rather important
for a good deal more than Rawls, so it may be of interest.
Rawls’ theory [if one leaves to one side the baroque
elaborations that fill most of the book, as I shall], turns on the notion that
I have called the “inequality surplus.”
The baseline payoff for practices chosen by the rationally self-interested
individuals in the bargaining game at the heart of the theory is an equal distribution
of rewards – money, or goods, or, in Rawls’ rather odd final version, an Index
of Primary Goods. But inequalities will
be allowed, indeed will be chosen by the bargainers, so long as they work to
the benefit of everyone [in the original version] or to the benefit of the
least advantaged representative individual [in the final version.] I am aware that in the final version it is
not the arrangement of practices, like factories or universities, but rather the
arrangement of the entire society, that is at issue, but that notion is so unhinged
from the original idea of the bargaining game that it makes no sense at all, so
forget it.
The core idea in Rawls’ theory is that it may be necessary to
pay certain individuals more to induce them to fill roles for which they are
especially well suited and in the filling of which they will increase the total
output of the practice. If the unequal
extra pay needed to draw them into these roles is less than the total extra
output of the practice with them in those roles, then there will be an inequality surplus that can be spread
around to make everyone [or the least advantaged representative individual]
better off. And since they are by hypothesis
not envious [this is the point of that ad
hoc assumption], they will prefer the practice with the unequal rewards.
OK, got it?
Now let’s try to imagine how this is supposed to work out. Let us assume the practice in
question is a firm that buys semi-finished components of widgets and assembles
them for shipment to wholesalers.
Suppose the firm has 100 employees:
one CEO, four middle managers, ninety assembly line workers, four
loading dock loaders and unloaders, and one cleaning person who mops the
floors, empties the waste baskets, and cleans the toilets. Let us assume that with no differentiation in
pay, all one hundred employees earn the same annual wage: $30,000.
In comes a time and motion expert who studies the factory
and gives aptitude tests to the employees.
He then reports that if someone especially talented at corporate
management is tapped for the job of CEO, which is currently filled by random
assignment, the output of the factory can be sufficiently improved so that
every employee can make $40,000 a year.
[You will notice that this is a philosophical argument, which is to say
no effort at all has been made to achieve realism.] Everyone
prefers that state of affairs, of course, because they all have positive
marginal utility for money, or an Index of Primary Goods, or whatever. More generally, if each person in the firm is
assigned to the job for which he or she is best suited, the resulting increase in
firm income can be spread around equally so that everyone prefers the new
system of assignments to the old random assignments. If some jobs require special training, the
costs of that training can be socialized, paid for by the firm as a whole. Everyone is better off. What’s not to love?
But there is a problem [there has to be a problem of this
sort, because otherwise Rawls’ entire theory is unnecessary.] The time and motion
expert reports that the person best suited to fill the position of CEO is
Genevieve, the cleaning lady. And
Genevieve does not want to be CEO. She
wants to keep her mop and pail and sponges.
“But you will make an extra $10,000 a year if you are running the
company. So will everyone else, but that
does not concern you, of course, because you are rationally self-interested. But surely you prefer the extra money,
because you have positive marginal utility for an Index of Primary Goods, and
in this case the only Primary
Good at issue is money.”
“What will I have to do to prepare myself to run the
company?”
“Well, first you will have to get a BA, which we will pay
for and then you will have to get an MBA, which we will also pay for.”
“So I am supposed to give up cleaning toilets to read
Shakespeare and Austen and Rousseau and Marx, and study microeconomics, and then
I have to quit mopping the floors and sit in a corner office and give orders,
and all for a lousy ten thousand extra?
No thank you. I prefer dusting
the machinery and emptying the waste baskets.”
“Suppose we pay you $139,000 instead of $40,000 and get the
extra $99,000 by reducing the other workers to $39,000 each? They will still prefer that to the $30,000
they are earning now.”
“Well, all right, but no art classes. I draw the line at that!”
I know this sounds silly, but think about it. Rawls’ theory depends completely on some
situation like this, and so does all of
modern economics and sociology! The
ideological rationalization for economic inequality is that the top jobs are so
unpleasant that talented people have to be bribed to accept them.
But that is just nonsense.
13 comments:
Perhaps I miss the point but it seems to me the notion of "well suited" has to be further defined. As it stands (or at least as far as we are told), Genevieve has no advanced skills and has little interest in acquiring such skills. How is she evaluated as best suited?
It doesn't matter!!! You can add anything you wish to the thought experiment. Indeed, the more eager she is to be CEO, the less reason there is to pay her extra, right?
It seems to me that a Rawlsian "super manager" is one who, by the luck of the genetic draw, is above average in intelligence and managerial ability. But such a person may not want to put forth the effort to get the education required. (I'm from a blue collar family and went to blue collar schools; many of my high school classmates, who had the ability, simply didn't want to go to college; they'd had enough of school.) Or, even with qualifications, reasonable people may not want the added responsibility and headaches involved with the top positions. I believe there are many academics who don't want to be department chairs. Law schools, I'm given to understand, sometimes have a hard time talking faculty into giving up their teaching and research to become dean and spend most of their time that is not taken up with administrative trivia schmoozing alumns for donations. It takes a lot of bucks to induce them.
So the added salary they would get simply by raising the total output and dividing it equally would be insufficient to entice them. And of course, if no one wanted to be the cleaning person, that job would need to have an extra salary boost too. And is the big salary bump for professors needed to woo gifted academics away from the enticing job of department secretary? Really?
I mean, think about it seriously for a moment. But notice, Rawls' argument presupposes that the top jobs are fundamentally unappealing, and require not just higher salaries [that follows from the fact that in those jobs they increase total output] but extra high salaries to get them to do the tedious work of running a factory rather than the enticing work on the production line.
It would be fun to offer professors the choice of being departmental secretaries and see how many would choose that line of work [even though the job of secretary has the attraction of working 48 weeks a year 5 days a week rather than 32 weeks a year 3 days a week.]
They might not work for a monetary reward, but would get a reward in status and/or intangible rewards which function the same way as money and would grease the greed so to speak
One of the weaker aspects of Understanding Rawls was your decision to read Rawls’ argument for the difference principle in A Theory of Justice primarily in relation to the 1958 “Justice as Fairness” article. And while it may have seemed germane in the late seventies, it would now be highly idiosyncratic to do so. The final statement of the argument, of course, is the one presented in 2001’s Justice as Fairness: A Restatement.
I would also suggest, that it is clear (in light of the full Rawlsian corpus) that Rawls does not think the use of the maximin rule for choice under uncertainty is a knock-down argument for the difference principle; in part I of A Theory of Justice, the appeal to the maximin rule is used to vindicate the choice of a conception of justice that accepts the the priority of liberty (the two principles of justice) over those that do not (classical utilitarianism and average utilitarianism).
Rawls came to see, and state more explicitly, that the primary competitor of justice as fairness (as he saw it) was not utilitarianism per se, but so-called “mixed conceptions” which accept the priority of liberty (or some suitable version of the first principle) but replace the difference principle with some form of restricted average utilitarianism. In chapter V of A Theory of Justice, Rawls argues that mixed conceptions, since they must set some social minimum, must rely on ad hoc considerations in order to do so; and, whatever principled basis might be used to set a social minimum might be just an unconscious application of the difference principle. In Justice as Fairness: A Restatement, Rawls indicates that he is less impressed with this argument, but introduces considerations of publicity, reciprocity, and stability to argue that the difference principle would be chosen as the second principle rather than a principle of restricted average utility.
A further question turns on to what degree the difference principle is in fact a justification of inequality at all. On this score, chapter 4 of G. A. Cohen’s Rescuing Justice and Equality (“The Difference Principle”) is worth consulting. (The whole first half of the book, really.)
I think the cleaning woman/CEO or department secretary/department chair comparison is not the issue. Presumably, the cleaning woman and the secretary aren't qualified to do the other jobs. They likely would prefer those other jobs, and so do the people who already have them,but since the pay is the same, it's a job enjoyment issue, not an economic one.
When I read that part of Rawls, the examples that came to my mind were the ones I mentioned, Department Chair, Law School Dean. Many academics who are qualified for those positions are perfectly happy staying where they are rather than take them. These are the people I think Rawls is talking about, not those like the cleaning woman and the secretary who aren't qualified to be chair or dean.
As an addendum to my previous remark:
"Despite the formal resemblance between the difference principle as a principle of distributive justice and the maximin rule as a rule of thumb for decisions under uncertainty... the reasoning for the difference principle does not rely on this rule. The formal resemblance is misleading." Justice as Fairness: A Restatement, 94-95.
How about the brain (or in our case back) surgeon who is willing to be on call 24 hours a day, 7 days a week, which means that he or she can never get very drunk or very stoned, who has gone through a very rigorous training, etc.? Couldn't we pay the surgeon a bit more than we pay your ordinary worker like myself?
Well, as I said, he is the world's leading authority on what he thinks, so I take back my claim that he remained wedded to the idea. A pity. It is what was interesting in the argument.
Alternatively, the ideological rationalization for economic inequality is that everybody gets what s/he deserves?
Dean, that is indeed the neoclassical economic rationale. Needless to say, it is not true, even in the perfect world of their theorizing.
Prof. Wolff,
Isn’t the implication of your critique, then, that Rawls’ theory of justice is far more egalitarian than most people assume it to be? After all, if no one needs a monetary incentive to take such a job since it’s already inherently desirable, then doesn’t that mean that the Difference Principle wouldn’t license such an inequality in the first place?
Post a Comment