TW ELVE

Is Wealth a Value?

I n t h i s e s s a y I consider and reject a political theory about law often called

the economic analysis of law. (That name is the title of an extended book by
Richard Posner,1 and I shall be concerned largely, though not entirely, with
arguments that Posner has himself presented.) The economic analysis of law
has a descriptive and a normative limb. It argues that common law judges,
at least, have on the whole decided hard cases to maximize social wealth,
and that they ought to decide such cases in that way. I shall discuss the nor-
mative limb of the theory mainly, although at the end of the essay I shall
argue that the normative failures of the theory are so great that they cast
doubt on its descriptive claims, unless these descriptive claims can be em-
bedded within a very different normative theory.
The concept of wealth maximization is at the center of both the descrip-
tive and normative aspects of the theory. But it is a concept that is easily
misunderstood, and it has been misunderstood, in a certain way, by its
critics. “Wealth maximization” is a term of art within the theory, and is
not intended to describe the same thing as “Pareto efficiency.” In this
introductory section, I shall try to explain each of these terms, to show why
it misunderstands the economic analysis of law to suppose, as critics have,
that the lawyer’s definition of the former is a botched attempt to capture
the meaning of the latter.
Wealth maximization, as defined, is achieved when goods and other re-
sources are in the hands of those who value them most, and someone values
a good more only if he is both willing and able to pay more in money (or in
the equivalent of money) to have it. An individual maximizes his own
wealth when he increases the value of the resources he owns; whenever he
is able, for example, to purchase something he values for any sum less than
the most he would be willing to pay for it. Its value to him is measured by
the money he would pay if necessary; if he is able to pay, say $4, for what he
would pay $5 to have if necessary, his wealth has been increased by $1. So-
ciety maximizes its wealth when all the resources of that society are so dis-
tributed that the sum of all such individual valuations is as high as possible.
238 The Economic View o f Law

There are many conceptual difficulties in this idea of individual and so-
cial wealth maximization. Some of these will emerge in the course of our
discussion, but one is sufficiently isolable that it can be disposed of now. For
most people there is a difference between the sum they would be willing to
pay for something that they do not have and the sum they would take in
exchange for it if they already had it. Sometimes the former sum is
greater— the familiar “grass is greener” phenomenon that leads someone to
covet his neighbor’s property more than if it were his own. If many people
were often in that position, then social wealth maximization would be in-
herently unstable. Social wealth would be improved by a transfer of some
property from A to B, but then improved by a retransfer from B to A, and so
on. In these circumstances, that is, wealth maximization would be a cyclic
standard— a very disagreeable property in a standard of social improve-
ment. The second case is perhaps more common (although not either more
or less rational): someone will ask more for something he owns than he
would pay to acquire it. When I am lucky enough to be able to buy Wim-
bledon tickets in the annual lottery for <£5,1 will not sell them for, say, £ 5 0,
although I will certainly not pay £ 2 0 to buy them when I lose in the lottery.
If many people are in that position with respect to many goods, then wealth
maximization will not be path-independent; the final distribution that
achieves a wealth maximization will be different, even given the same ini-
tial distribution, depending upon the order in which intermediate transfers
are made. Path-dependency is not so serious a flaw as cyclicity, but does
nevertheless introduce an element of arbitrariness into any scheme of trans-
fers designed to promote social wealth maximization.
Neither Posner nor other proponents of economic analysis of law seem
much bothered by either possibility. They assume, perhaps, stipulations of
rationality that preclude differences in pay-or-take value of this sort. Or
perhaps they are concerned principally with the behavior of commercial
firms where such stipulations would not seem so arbitrary. It will do no
harm, however, to tighten their definitions. We may say that the goal of
wealth maximization is served by a particular transfer or distribution only
when that transfer would increase social wealth measured by what the per-
son into whose hands the good falls would pay if necessary to acquire it, and
also by what he would take to part with it. In cases where the two tests dis-
agree, the standard of social wealth maximization is indeterminate. Indeter-
minacy in some cases is no great objection to any standard for social
improvement, provided that such cases are not disagreeably numerous.
The familiar economist’s concept of Pareto efficiency (or Pareto optimal-
ity) is a very different matter. A distribution of resources is Pareto efficient
if no change in that distribution can be made that leaves no one worse off
and at least one person better off. It has often been pointed out that almost
any widespread distribution of resources meets that test. Even willing
 Is Wealth a Value? 239

trades that improve the position of both parties may adversely affect some
third party by, for example, changing prices. It would be absurd to say that
judges should make no decision save those that move society from a Pareto-
inefficient to a Pareto-efficient state. That constraint is too strong, because
there are few Pareto-inefficient states; but it is also too weak because, if a
Pareto-inefficient situation does exist, any number of different changes
would reach a Pareto-efficient situation and the constraint would not
choose among these.
Suppose no court has decided, for example, whether a candy manufac-
turer is liable to a doctor if the manufacturer’s machine makes it more dif-
ficult to practice medicine in an adjacent building.2 The doctor does not
have a recognized legal right to damages or an injunction, but neither does
the manufacturer have a recognized right to run his machine without pay-
ing such damages. The doctor sues the candymaker, and the court must de-
cide which of these two rights to recognize. Neither decision will be
Pareto-superior to the situation before the decision, for either decision will
improve the position of one party at the expense of the other. Both deci-
sions will reach a Pareto-efficient result, for no further change in the legal
position would benefit one without hurting the other. So the requirement,
that the court should decide in favor of a Pareto-superior rule, if one is
available, would be useless in such a case.
But the different advice, that the court should choose the rule that maxi-
mizes social wealth, is far from useless. R. H. Coase has argued that, if trans-
action costs were zero, it would make no difference to that goal which of
the two decisions the court made.3 If the decision did not in itself maximize
wealth, then the parties would negotiate a solution that did. But since trans-
action costs are always positive, it will in practice make a difference. If the
candymaker would lose $10 by not running his machine, and the doctor
would lose only $9 if the machine were run, then social wealth would not be
maximizedJ3y a rule giving the doctor a right to prevent the running of the
machine, if transaction costs would exceed $1. The judge should, therefore,
choose so that goods (in this case the right to practice medicine free from
noise or the right to make candy free from injunction) are given directly, by
his decision, to that party who would purchase the right if not assigned it,
and would not sell it if assigned it, in both cases assuming that transaction
costs were zero. In many cases this requirement, unlike the requirement of
Pareto superiority, would dictate a unique solution. If the candy manufac-
turer makes enough through his noisy machine to compensate fully the
doctor for his lost practice and still have profit left over, as he does on the
figures just assumed, then the right to make the noise without compensation
should be assigned to the candymaker. Of course, that will not produce the
distribution that would have been achieved if the right had been assigned to
the doctor and there were no transaction costs. In that case the doctor
240 The Economic View o f Law

would have had something over $9 and the candymaker something less than
$1. Now the candymaker will have $10 and the doctor nothing. But that
produces more total social wealth than the only actual alternative, given
the transaction cost, which is that the candymaker have nothing and the
doctor $9.
So the theory of wealth maximization is both different from the theory of
Pareto efficiency and more practical. The economic analysis of law, which
makes the concept of wealth maximization central, must therefore be dis-
tinguished from the economists’ analysis of law, that is, from the application
to legal contexts of the economists’ notion of efficiency, which is Pareto effi-
ciency. When an economist asks whether a rule of law is efficient, he usually
means to ask whether the situation produced by the rule is Pareto-efficient,
not whether it is wealth maximizing. Much confusion could have been
avoided if Posner and others had not used the words “economic” or “effi-
cient” in their description of their own work. Economists would not have
been so concerned to point out that these words are obviously not used in
their normal professional sense. They would not then have supposed that
Posner and his colleagues had made some simple conceptual mistakes.

B u t n o w c o m e s the nerve of the problem. Economic analysis holds, on its

normative side, that social wealth maximization is a worthy goal so that ju-
dicial decisions should try to maximize social wealth, for example, by as-
signing rights to those who would purchase them but for transaction costs.
But it is unclear why social wealth is a worthy goal. Who would think that a
society that has more wealth, as defined, is either better or better off than
a society that has less, except someone who made the mistake of personi-
fying society, and therefore thought that a society is better off with more
wealth in just the way any individual is? Why should anyone who has not
made this mistake think social wealth maximization a worthy goal?
There are several possible answers to this question, and I shall start by
deploying a number of distinctions among them. (1) Social wealth may be
thought to be itself a component of social value— that is, something worth
having for its own sake. There are two versions of this claim, (a) The im-
modest version holds that social wealth is the only component of social
value. It argues that the only respect in which one society may be better or
better off than another is that it may have more social wealth, (b) The mod-
est version argues that social wealth is one component of social value
among others. One society is pro tanto better than another if it has more
wealth, but it might be worse overall when other components of value, in-
cluding distributional components, are taken into account.
(2) Social wealth may be thought to be, not a component, but an instru-
ment of value. Improvements in social wealth are not valuable in them-
 Is Wealth a Value? 241

selves, but valuable because they may or will produce other improvements
that are valuable in themselves. Once again, we may distinguish different
versions of the instrumental claim, (a) The causal claim argues that im-
provements in social wealth themselves cause other improvements: im-
provements in wealth, for example, improve the position of the worst-off
group in society by alleviating poverty through some invisible hand process,
(b) A second claim argues that improvements in social wealth are ingre-
dients of social value, because although they do not work automatically to
cause other improvements, they provide the material for such improve-
ments. If a society has more wealth, it is better off because it is in a position
to use that increased wealth to reduce poverty, (c) A third claim holds that
social wealth is neither a cause nor an ingredient of social value, but a sur-
rogate for it. If society aims directly at some improvement in value, such as
trying to increase overall happiness among its members, it will fail to pro-
duce as much of that goal than if it instead aimed at improving social
wealth. Social wealth is, on this “false-target” account, a second-best goal,
valued not for its own sake, nor because it will cause or can be used to bring
about other improvements, but because there is a sufficiently high correla-
tion between improvements in social wealth and such other improvements
to make the false target a good target.
Another distinction cuts across these. Each of these modes of social
wealth claims, except the immodest version of the component-of-value
claim, may be combined with some functional claim of institutional respon-
sibility which argues that it is the special function of courts to pursue social
wealth single-mindedly, although it is not necessarily the function of, for
instance, legislatures to do so. It might be said, for example, that although
wealth maximization is only one among several components of social value,
it is nevertheless a component that courts should be asked single-mindedly
to pursue, leaving other components to other institutions. Or that although
social wealth is only an ingredient of social value it should be left to courts
to maximize that ingredient, on the understanding that the further use of
the ingredient is the province of other institutions. Or that social wealth is a
value surrogate for courts, because courts cannot for some reason pursue
the true target directly, though other institutions can and therefore need no
surrogate or perhaps need a different surrogate. I shall call a theory of this
sort a strong institutional theory— “institutional” because it specifies rea-
sons why one institution should pursue social wealth maximization, and
“strong” because it requires that those institutions do so single-mindedly.
The normative claim of economic analysis, then, admits of many varia-
tions. Calabresi, Posner, and other advocates of that analysis have not been
as clear as they might be about which variation they wish to promote, so
any thorough discussion of their claims must consider different possibilities
and paint on a reasonably wide canvas. I shall begin by considering whether
242 The Economic View of Law

the claim that social wealth is a component of value, in either the immodest
or the modest versions of that claim, is a defensible idea.

I t h i n k it is plain it is not. Perhaps no one thinks it is, although there has

been much careless rhetoric on this score.4 Before I provide an illustration
that seems to me decisive against the component-of-value theory, however,
I shall try to clarify the point at issue. If economic analysis argues that law-
suits should be decided to increase social wealth, defined in the particular
way described, then it must show why a society with more wealth is, for
that reason alone, better or better off than a society with less. I have distin-
guished, and now propose to consider, one form of answer: social wealth is
in itself a component of value. That answer states a theory of value. It holds
that if society changes so that there is more wealth then that change is in
itself, at least pro tanto, an improvement in value, even if there is no other
change that is also an improvement in value, and even if the change is in
other ways a fall in value. The present question is not whether a society that
follows the economic analysis of law will produce changes that are im-
provements in wealth with nothing else to recommend them. The question
is whether such a change would be an improvement in value. That is a
question of moral philosophy, in its broadest sense, not of how economic
analysis works in practice. If the answer to my question is no— a bare im-
provement in social wealth is not an improvement in value— the claim that
social wealth is a component of value fails, and the normative claim of eco-
nomic analysis needs other support.
Consider this hypothetical example. Derek has a book Amartya wants.
Derek would sell the book to Amartya for $2 and Amartya would pay $3 for
it. T (the tyrant in charge) takes the book from Derek and gives it to Amart-
ya with less waste in money or its equivalent than would be consumed in
transaction costs if the two were to haggle over the distribution of the $1
surplus value. The forced transfer from Derek to Amartya produces a gain
in social wealth, even though Derek has lost something he values with no
compensation. Let us call the situation before the forced transfer takes
place “Society 1” and the situation after it takes place “Society 2.” Is So-
ciety 2 in any respect superior to Society 1? I do not mean whether the gain
in wealth is overridden by the cost in justice, or in equal treatment, or in
anything else, but whether the gain in wealth is, considered in itself, any
gain at all. I should say, and I think most people would agree, that Society 2
is not better in any respect.5
It may be objected that in practice social wealth would be maximized by
rules of law that forbid theft and insist on a market exchange, when it is
feasible, as it is in my imaginary case. It is true that Posner and others rec-
ommend market transactions except in cases in which the transaction costs
 Is Wealth a Value? 243

(the costs of the parties identifying each other and concluding an agree-
ment) are high. But it is crucial that they recommend market transactions
for their evidentiary value. If two parties conclude a bargain at a certain
price, we can be sure that wealth has been increased (setting aside problems
of externalities), because each has something he would rather have than
what he gave up. If transaction costs are “high” or a transaction is, in the
nature of the case, impossible, Posner and others recommend what they call
“mimicking” the market, which means imposing the result they believe a
market would have reached. They concede, therefore, or rather insist, that
information about what parties would have done in a market transaction
can be obtained in the absence of the transaction, and that such information
can be sufficiently reliable to act on.
I assume, therefore, that we have that information in the book case. We
know that there will be a gain in social wealth if we transfer the book from
Derek to Amartya. We know there will be less gain (because of what either
or both might otherwise produce) if we allow them to “waste” time hag-
gling. We know there can be no more gain in social wealth if we force
Amartya to pay anything to Derek in compensation. (Each would pay the
same in money for money.) If we think that Society 2 is in no respect supe-
rior to Society 1, we cannot think that social wealth is a component of
value.
It may now be objected, however, that wealth maximization is best
served by a legal system that assigns rights to particular people, and then
insists that no one lose what he has a right to have except through a volun-
tary transaction. Or (if his property has been damaged) in return for appro-
priate compensation ideally measured by what he would have taken for it in
such a transaction. That explains why someone who believes that wealth
maximization is a component of value may nevertheless deny that Society 2
is in any way better than Society 1. If we assume that Derek has a right to
the book under a system of rights calculated to maximize wealth, then it of-
fends, rather than serves, wealth maximization to take the book with no
compensation.
I shall discuss later the theory of rights that is supposedly derived from
the goal of maximizing wealth. We must notice now, however, that the goal
justifies only instrumentally rights, like Derek’s right to the book. The insti-
tution of rights, and particular allocations of rights, are justified only insofar
as they promote social wealth more effectively than other institutions or al-
locations. The argument for these rights is formally similar to the familiar
rule-utilitarian account of rights. Sometimes an act that violates what most
people think are rights— such as taking Derek’s book for Amartya— im-
proves total utility. Some rule utilitarians argue that such rights should nev-
ertheless be respected, as a strategy to gain long-term utility, even though
utility is lost in any isolated case considered by itself.
244 The Economic View o f Law

This form of argument is not to the point here. I did not ask whether it is
a wise strategy, from the standpoint of maximizing social wealth in the long
run, to allow tyrants to take things that belong to one person and give them
to others. I asked whether, in the story of Amartya and Derek, Society 2 is
in any respect superior to Society 1. The utilitarian, assuming that Amartya
would get more utility than Derek would lose, might reply that it is. He
might say that, if we confine our attention only to this case, Society 2 is in
every way better because there is more happiness, or less suffering, or what-
ever. He would add, however, that we should nevertheless impose on the
tyrant a rule forbidding the transfer, because, although the act makes
the immediate situation better, its consequences will make the situation
in the future much worse. This distinction is important, because a utilitar-
ian who takes this line must concede that, if the tyrant’s act would not have
the long-term adverse utility consequences he supposes (because the act
could be kept secret, or because a suitably limited exception to the general
rule he endorses could be carved out and maintained), then the tyrant
should so act. Even if the utilitarian insists that a rule forbidding the trans-
fer in all cases will improve long-term utility, he still concedes that some-
thing of value is lost through the rule, namely, the utility that would have
been gained but for the rule.
The wealth maximizer’s answer to my question about Amartya and
Derek— that economic analysis would not recommend a set of legal rules
permitting the tyrant to transfer the book without compensation— is simply
an evasion. Like the reply that market exchanges provide the most reliable
information about value, it misunderstands the force of my story. I still ask
whether the situation is in any respect better if the transfer is made. If So-
ciety 2 is not in any way superior to Society 1, considered in themselves,
then social wealth is not even one among several components of social
value.
I have assumed so far, however, that you will agree with me that Society
2 is not superior. Perhaps I am wrong. You may wish to say that a situation
is better, pro tanto, if goods are in the hands of those who would pay more
to have them. If you do, I suspect it is because you are making a further as-
sumption, which is this: if Derek would take only $2 for the book and
Amartya would pay $3, then the book will provide more satisfaction to
Amartya than it does to Derek. You assume, that is, that the transfer will
increase overall utility as well as wealth. But Posner, at least, is now explicit
that wealth is conceptually independent of utility. He now allows that in-
terpersonal comparisons of utility make sense and holds that increases in
wealth may produce decreases in utility and vice versa.6 (He relies on cases
in which this is so as part of his argument that economic analysis is superior
to utilitarianism as a moral theory.)
I must thus make my example more specific. Derek is poor and sick and
 Is Wealth a Value? 245

miserable, and the book is one of his few comforts. He is willing to sell it for
$2 only because he needs medicine. Amartya is rich and content. He is will-
ing to spend $3 for the book, which is a very small part of his wealth, on the
odd chance that he might someday read it, although he knows that he prob-
ably will not. If the tyrant makes the transfer with no compensation, total
utility will sharply fall. But wealth, as specifically defined, will improve. I
do not ask whether you would approve the tyrant’s act. I ask whether, if the
tyrant acts, the situation will be in any way an improvement. I believe it
will not. In such circumstances, the fact that goods are in the hands of those
who would pay more to have them is as morally irrelevant as the book’s
being in the hands of the alphabetically prior party.
Once social wealth is divorced from utility, at least, it loses all plausibility
as a component of value. It loses even the spurious appeal given to utilitari-
anism by the personification of society. It is sometimes argued by utilitari-
ans that, since an individual is necessarily better off if he has more total
happiness in his entire life, even though less on many particular days, so a
society must be better off if it has more total happiness distributed across its
members even though many of these members have less. That is, I think, a
bad argument in two different ways. First, it is not true that an individual is
necessarily better off if he has more total happiness over his life without re-
gard to distribution. Someone might well prefer a life with less total plea-
sure than a life of misery with one incredibly ecstatic month, and perjured
Clarence would not have relived the agony of his dream “Though ’twere to
buy a world of happy days.” ' Second, society is not related to individual cit-
izens as an individual is related to the days of his life. The analogy is, there-
fore, one way of committing the ambiguous sin of “not taking seriously the
difference between people.”
The parallel argument on behalf of social wealth maximization is, how-
ever, much worse. It is false that even an individual is necessarily better off
if he has more wealth, once having more wealth is taken to be independent
of utility information. Posner concedes that improvements in wealth do not
necessarily lead to improvements in happiness. He should also concede that
they sometimes lead to a loss in happiness because, as he says, people want
things other than wealth, and these further preferences may be jeopardized
by more wealth. That is, after all, a staple claim of sentimental fiction and
quite unsentimental fairy tales. Suppose, therefore, that an individual faces
a choice between a life that will make him happier (or more fulfilled, or
more successful in his own lights, or whatever) and a life that will make him
wealthier in money or the equivalent of money. It would be irrational of
him to choose the latter. Nor, and this is the crux, does he lose or sacrifice
anything of value in choosing the former. It is not that he should, on bal-
ance, prefer the former, recognizing that in the choice he sacrifices some-
thing of value in the latter. Money or its equivalent is useful so far as it
246 The Economic View o f Law

enables someone to lead a more valuable, successful, happier, or more

moral life. Anyone who counts it for more than that is a fetishist of little
green paper.

I t is i m po r t a n t to notice that the Derek-Amartya story shows the failure

not only of the immodest but also of the modest version of the theory that
social wealth is a component of value. For the story shows not merely that a
gain in wealth may be outweighed by losses in utility or fairness or some-
thing else. It shows that a gain in social wealth, considered just in itself and
apart from its costs or other good or bad consequences, is no gain at all.
That denies the modest as well as the immodest theory. I shall therefore
take this opportunity to comment on a familiar idea that, on its most plau-
sible interpretation, presupposes the modest theory, that is, that social
wealth is one among other components of social value.
This is the idea that justice and social wealth may sensibly be traded off
against each other, making some sacrifice in one to achieve more of the
other. Guido Calabresi, for example, begins The Costs o f A ccidents by no-
ticing that accident law has two goals, which he describes as “justice” and
“cost reduction,” and notices also that these goals may sometimes conflict
so that a “political” choice is needed about which goal should be pursued.8
The same point is meant to be illustrated by the indifference curves I have
seen drawn on countless blackboards, on space defined by axes one of which
is labeled “justice” (or sometimes “morality”) and the other “social wealth”
(or sometimes “efficiency”).
Whose indifference curves are supposed to be drawn on that space? The
usual story speaks of the “political” or “collective” choice in which “we”
decide how much justice we are willing to give up for further wealth or
vice versa. The suggestion is that the curves represent individual choices (or
collective functions of individual choices) over alternative societies defined
as displaying different mixes of justice and wealth. But what sort of choice is
the individual, whose preferences are thus displayed, supposed to have
made? Is it a choice of the society in which he would like to live, or the
choice of the society he thinks best from the standpoint of morality or some
other normative perspective? We shall have to consider these two interpre-
tations in turn.
If the former, self-interest might be thought to enter directly in a way an-
tagonistic to justice, as in the case of an individual deciding whether to lead
a perfectly just life that will leave him poor, or a life in which he sometimes
acts unjustly but in which he is richer, or a life of many very unjust acts in
which he is richer still. Since I believe that people can (and often do) act in
a way they know is unjust, I acknowledge that individuals “trade off” justice
against personal welfare in their own lives. But what sense does it make to
 Is Wealth a Value? 247

suppose that they tradeoff justice against, not welfare in their own lives,
but wealth over the society as defined by economic analysis?
Perhaps the point is that an individual chooses a society that has more
rather than less wealth as a whole because the antecedent probability is that
he will have more wealth personally in a richer society. This makes the
supposed preferences something like those displayed in Rawls’s original po-
sition. Individuals choose a mix of justice and efficiency with an eye to max-
imizing their individual utility under conditions of dramatic uncertainty;
or, rather, trading off gains in their prospects, so conceived, against losses in
the just character of the society. (This is very different from the choice
made in Rawls’s own version of the original position, in which people maxi-
mize their antecedent self-interest not as some trade-off against justice, but
as part of a demonstration— by Rawls, not them— of what principles con-
stitute justice.)
Individuals in this exercise would be ill-advised to take gains in social
wealth as some index to gains in their own antecedent welfare, even under
conditions of uncertainty about the role they will occupy. Just under those
conditions, they will use a very different index. Which index they will use
will depend upon whether they decide to draft their preferences over so-
ciety in the language of utility or the language of wealth. Which language
they use— the language of utility or the language of individual wealth— will
depend upon calculations about which vocabulary will, in practice, maxi-
mize antecedent welfare. If they choose the language of utility, then, as
Hirsanyi and Mackie and others argue, they will choose, as the surrogate for
maximizing their own antecedent welfare, average utility. If they choose (as
I think they should) the language of individual wealth, they will certainly
not choose, as that surrogate, that function of individual wealth constituted
by social wealth as defined by the economic analysis of law. That would be
crazy. Nor will they choose, for that surrogate, average individual wealth,
because of the effects of marginal utility. They would be better advised to
choose something much closer to maximin of individual wealth, for exam-
ple, which is Rawls’s second principle. I do not think that they would
choose only maximin— they would allow some gains for those better off, if
sufficiently large, to outweigh small losses to those worse off. But if their
only choice were maximin or highest social wealth, they would certainly
choose the former.
But surely this is all irrelevant. Calabresi and others contemplate actual
political choices— they suppose that the economic analysis of law is useful
because it shows how much wealth is lost if some other value is chosen. But
in that case we cannot understand the axis of wealth or efficiency, in the in-
difference curves as generally offered, as a surrogate for judgments about
antecedent individual welfare under conditions of uncertainty. We must
understand the axis as representing judgments about individual welfare, to
248 The Economic View of Law

be traded off against justice, as things actually stand. No particular individ-

ual will, then, be concerned about social wealth (or, indeed, about Pareto
efficiency). It makes no sense for him to trade off anything, let alone justice,
for th at He will be concerned with his individual fate, and since, by hy-
pothesis, he now knows his actual position, he can choose among societies
by trading off justice against increases in his individual welfare in these dif-
ferent societies. Social wealth (or Pareto efficiency) plays no role in these
calculations.
Let us turn to the second interpretation of the supposed trade-off choice.
An individual is supposed to be choosing which mix of justice and wealth
represents, not the society in which he, as an individual with both moral
and self-interested motives, would prefer to live, but the morally best so-
ciety, all things considered. The very idea of a trade-off between justice and
wealth now becomes mysterious. If the individual is to choose the morally
best society, why should not its justice alone matter?
We might expect one of two replies to that question. It might be said,
first, that justice is not the only virtue of a good society. It surely makes
sense, from a normative perspective, to speak of the trade-off between jus-
tice and culture, and also to speak of the trade-off between justice and social
wealth, as two distinct, sometimes competing social virtues. The second
reply is different in form but similar in spirit. It suggests that, when people
speak of a trade-off between justice and social wealth, they use “justice” to
refer to only part of what that word means in ordinary language and in po-
litical philosophy— that is, they use it to refer to the distributional and
meritocratic or desert features of justice in the wider sense. They mean the
trade-off between those specific aspects of justice and other aspects that are
comprehended under “wealth maximization.”
These two replies are similar in spirit because they both assume that
wealth maximization is a component of social value. In the first, wealth
maximization is treated as a component competitive with justice and, in the
second, as a component of justice but competitive with other components
of that concept. Both replies fail, for that reason. It is absurd to consider
wealth maximization to be a component of value, within or without the
concept of justice. Remember Derek and Amartya.
Of course, if someone denies that wealth is a component of value, but
argues that it is sometimes instrumental in achieving value, in one of the
senses we distinguished earlier in this essay, he would not speak of a trade-
off between justice and wealth. Or rather he would be confused if he did. It
makes no sense to speak of trading off means against ends, or of people
being indifferent about different mixes of a particular means and the end it
is supposed to serve. Someone who speaks this way must have in mind an
entirely different point. He might mean, for example, that sometimes we
achieve more of the desired end if we aim only at what is (in this sense) a
 Is Wealth a Value? 249

means. That is the “false-target” instrumental theory I mentioned before

and will discuss later. It entirely distorts that theory to describe it as requir-
ing some trade-off between justice and anything else.
But suppose I was wrong to take the trade-off described in the familiar
indifference curves, or in texts like Calabresi’s, to be a matter of individual
preferences, or some collective function of individual preferences. Perhaps
the choice is meant to be the choice of society as a whole, conceived as a
composite entity. I think that the choice is mentally represented this way,
although not reflectively, by many of those who speak of trade-offs between
justice and wealth. They have a personified community in mind, as the ref-
erence of the “we” in the proposition that “we” want a society of such-
and-such sort. That picture must be disowned when made explicit. It is a
silly and malign personification.
Even if society is personified in this silly way, it remains mysterious why
society so conceived would want a trade-off between justice and wealth.
First, the choice of wealth, taken to be independent of utility information,
would make no more sense for society as a composite person than it does for
individuals as actual people. Second, and more interesting, the reference of
“justice” would be lost. Justice (at least when the trade-off is in question) is
a matter of distribution— of the relation among individuals who make up
the society, or between the society as a whole and these individuals. Once
we personify the society so as to make the social choice an individual
choice, there is no longer anything to be considered under the aspect of jus-
tice. Society personified can, of course, still be concerned about questions or
ordering or distribution among its members. But the dimensions of such or-
derings do not include that of justice. An individual cares about the distri-
bution of benefits or experiences over the days of his life. But he does not
care under the aspect of justice.
None of these interpretations of the trade-off between justice and wealth
makes sense. I hope the idea, however familiar, soon disappears from eco-
nomic and political theory. My present point is more basic. The argument
thus far is as destructive of modest normative claims for economic analysis,
such as those Calabresi suggests, as it is of the most full-blown immodest
claims of Posner.

I n o w t u r n to the claim that a society with more wealth is better because

wealth bears some important instrumental connection— whether as cause,
as ingredient, or as false-target— to some independent component of value.
I characterized certain versions of the instrumental claim as “strong,” and
we must be careful to distinguish these from weaker claims. A weak instru-
mental claim argues merely that sometimes improvements in social wealth
cause improvements of other sorts. That is plainly sometimes so, for a vari-
250 The Economic View of Law

ety of reasons. If, for example, judges are able to increase wealth dramati-
cally by some decision they reach, then in, perhaps, a quarter-century ev-
eryone then alive may be better off than he would have been if the gain had
not been made, either because the increased wealth will be distributed by
political action so that even the poor benefit, or because the same result is
reached by some invisible hand mechanism with no direct political action.
But the weak instrumental claim— that sometimes this will be so— is insuffi-
cient to argue that judges should accept wealth maximization as the single
test for change in the common law, or even in some particular branch or
division of the common law. That argument requires the strong thesis that
judges who accept such a single test will produce more of what is indepen-
dently valuable, like the amelioration of poverty, than if they were to adopt
a more discriminating test and try to maximize wealth only in those cases in
which they have some special reason to think that they would thereby in-
crease the independent value.
This is an important point. The difference between a strong and a weak
instrumental claim is not only measured in scope. A strong theory need not
claim that judges must pursue wealth maximization as the only standard of
their decisions in all cases at law, or even in all common law cases or all tort
cases— although the more scope the claim has, the more interesting it is.
But the theory must claim that judges should pursue wealth single-mind-
edly over some class of cases specified independently of the instrumental
claim itself— that is, specified other than as “the cases in which maximizing
wealth will in fact produce the true goal.” If the normative limb of eco-
nomic analysis does not include at least some strong instrumental claim of
that sort— if it rests only on the weak and unelaborated claim that some-
times pursuing wealth will lead to other good results— then the normative
limb of the theory is boring and misleading: boring because no one will dis-
pute the claim, and misleading because the theory should then be named,
not after wealth, but after the so-far unspecified true goal that wealth is
taken sometimes to serve.
I shall assume, therefore, that if economic analysis rejects wealth as a
component of value and argues only that wealth maximization is instru-
mental toward some other conceptually independent goal or value, it
argues for that instrumental connection in some strong form, though I shall
not assume that the strong claim it makes has any particular scope. The
strong thesis need not suppose (of course, it need not deny) that in every
case a judicial decision that maximizes social wealth will improve the true
goal. But it must show why, if in some cases wealth maximization will not
have that desirable effect, it is nevertheless wise strategy to pursue wealth
maximization in all cases within the scope of the claim.
Any strong claim, even of limited scope, must specify the independent
goal or value that it supposes is advanced instrumentally by maximizing so-
cial wealth. Supporters of economic analysis might have any number of in-
 Is Wealth a Value? 251

dependent values in mind, or some structured or intuitionistic mix of differ-

ent independent values. We cannot test the instrumental claim for wealth
maximization until the independent value or mix of these is at least roughly
specified.
It is surprising that, in spite of the supposed popularity of economic
analysis, there have been few attempts to do this. This failure supports my
view that many lawyers have uncritically assumed that wealth is at least a
component of value. But in a recent article, and much more clearly in re-
marks prepared for a recent conference, Posner suggests different instru-
mental claims that he, at least, might be tempted to make.9 He suggests that
wealth maximization is a value because a society that takes wealth maximi-
zation to be its central standard for political decisions will develop other
attractive features. In particular, it will honor individual rights, encourage
and reward of variety of “Protestant” virtues, and give point and effect to
the impulses of people to create benefits for each other. Posner believes that
it will do better in promoting these attractive traits and consequences than
a society that takes, as its central standard for political decisions, either util-
itarianism or some “Kantian” position.10
The argument has the form of a strong instrumentalist claim of the causal
variety. It has very wide scope. It specifies a set of features of society— indi-
vidual rights, agreeable virtues, and humane instincts— that can plausibly
be taken to be components of value. It then suggests that the “right” mix of
these will be best obtained by a single-minded attention to wealth maximi-
zation as a standard for political decisions, including judicial decisions. The
trouble begins, however, when we ask what arguments he might offer to
support this strong and wide instrumentalist claim.
We may begin with the claim that wealth maximization will encourage
respect for individual rights. A society that sets out to maximize social
wealth will require some assignment of rights to property, labor, and so
forth. That,is a conceptual requirement, because wealth is measured by
what people are willing to pay, in money or its equivalent, but no one can
pay what he does not own, or borrow if he has nothing to pledge or if others
have nothing to lend. Society bent on maximizing wealth must specify what
rights people have to money, labor, or other property so that it can be de-
termined what is theirs to spend and, in this way, where wealth is im-
proved. A society is, however, not a better society just because it specifies
that certain people are entitled to certain things. Witness South Africa.
Everything depends on which rights society recognizes, and on whether
those rights should be recognized according to some independent test. It
cannot, that is, provide an instrumental claim for wealth maximization that
it leads to the recognition of certain individual rights, if all that can be said,
in favor of the moral value of these rights, is that these are the rights that a
system of wealth maximization would recognize.
There is, however, a danger that Posner’s argument will become circular
252 The Economic View o f Law

in that way. According to the economic analysis of law, rights should be as-
signed instrumentally, in such a way that the assignment of rights will ad-
vance wealth maximization. That is the principal use of the standard of
wealth maximization in the judicial context. Recall the case of the doctor
and the candymaker. The question put to the court was whether the doctor
should be recognized to have the right to stop the noisy machine. Economic
analysis does not suppose that there is some independent moral argument in
favor of giving or withholding that right. So it cannot be claimed, in favor of
economic analysis, that it points to what is independently, on moral
grounds, the right answer. On the contrary, it claims that the right answer is
right only because the answer increases social wealth.
Nor does Posner limit the scope of that argument— that assignments of
rights must be made instrumentally— to what might be called less impor-
tant rights, like the right to an injunction in nuisance or to damages in neg-
ligence. On the contrary, he is explicit that the same test must be used in
determining the most fundamental human rights of citizens, including their
right to life and to control their own labor rather than be slaves to others.
He counts it an important virtue of wealth maximization that it explains
why people have those rights. But if wealth maximization is only to be an
instrumental value— and that is the hypothesis now being considered— then
there must be some independent moral claim for the rights that wealth
maximization recommends. These rights cannot have a moral claim on us
simply because recognizing those rights advances wealth.
Let us, therefore, suppose that Posner believes people have a right to
their own bodies, and to direct their own labor as they wish, for some inde-
pendent moral reason. Suppose he also argues that wealth maximization is
of instrumental value because a society that maximizes wealth will recog-
nize just those rights. There remains a serious conceptual difficulty. The ar-
gument supposes that a social order bent only on wealth maximization,
which makes no independent judgments about the fairness of distributions
of resources, will recognize the rights of the “natural” owner to his own
body and labor. That is true only if the assumption of those rights can be
justified by the wealth maximization test, which requires that if rights to the
“natural owner’s” body or labor are assigned to someone else, he will never-
theless be willing and able to purchase these rights, at least, if we assume no
transaction costs.
We cannot, however, speculate intelligibly about whether someone
would purchase the right to his own labor unless we make some assump-
tions about the distribution of wealth. Posner acknowledges this. Indeed, he
uses this example— someone’s ability to purchase the right to his own labor
if he is made a slave— to make the point that whether someone can pur-
chase that right depends on his and others’ wealth, and in particular how
large a share of that wealth is that right. He says that in such a case “eco-
 Is Wealth a Value? 253

nomic analysis does not predict a unique allocation of resources unless the
initial assignment of rights is specified/’11 If A is B ’s slave he may not be
able to buy back the right to his labor; although if he were not B would not
be able to buy that right from him. If economic analysis makes someone’s
initial right to his own labor depend upon whether he would purchase the
right if assigned to another, that right cannot be “derived” from economic
analysis unless we already know who initially has the right. This appears to
be a serious circle. We cannot specify an initial assignment of rights unless
we answer questions that cannot be answered unless an initial assignment of
rights is specified.
Can we break out of this circle? We might, for example, stipulate that we
are to ask our question about who would purchase what in a state of nature
when no one has any rights to anything. I assume that means not only that
no one already owns his own labor, but also that no one has any money, the
equivalent of money, or anything else. In that case, the question is without
meaning, or, if it has meaning, the answer is that no one would purchase
anything.
We might more plausibly stipulate that we are to ask the question now,
that is, at a moment when other rights, including wealth, are in place
(which does not preclude asking it again later when we suspect that a differ-
ent answer might be available). There is, perhaps, a determinate answer to
the question of who values the right more under these circumstances. In
order to test the claim— that wealth maximization would (determinately)
assign the right to labor to the “natural owner”— we suppose that the right
to the labor of a certain easily distinguished group of people (say those with
IQs over 120) is taken from them (perhaps by some anti-emancipation proc-
lamation) and assigned to others. The present wealth of those who have lost
these rights (as well as the present wealth of those who have gained them) is
not otherwise disturbed. Can we say that at least most of those who have
lost their rights would now repurchase them or would but for transaction
costs?
We must remind ourselves that willingness to purchase these rights sup-
poses ability to purchase them— the ability to pay what those who have the
rights would ask in the market. It may be— it would be for most people
today— impossible to repurchase the right to their labor, because the value
of that labor represents more than half of their present wealth. Could they
borrow, in the money market, the necessary hinds? Posner speaks to this
possibility. He says, “No doubt the inherent difficulties of borrowing against
human capital would defeat some efforts by the natural owner to buy back
the right of his labor . . . even from someone who did not really value it
more highly than he did— but that is simply a further reason for initially
vesting the right in the natural owner.”12 These “inherent difficulties” must
be transaction costs or other market imperfections, because Posner is very
254 The Economic View o f Law

strict about how economic analysis must understand the verb “to value.”
Someone values something more than someone else (and the system of eco-
nomic analysis depends on this) only if he is willing (and able) to pay more
for it. If (for reasons other than market imperfections) the natural owner is
unable to pay what the owner of the right would take, then he does not
value it more.
So let us assume that the “inherent difficulties” can be overcome, so that
someone who has lost the right to his labor can borrow against the dis-
counted value of his future labor. Will he thereby gain enough capital so
that we can be confident that he (or most people in his position) will be able
to purchase the right to his labor back from someone else? Almost cer-
tainly not, because the monetary value of his future labor is unlikely to be
worth more to him, for this purpose, than it is to someone else.
Suppose someone called Agatha who is poor but who can write detective
stories so brilliantly that the public will relish and pay for as many books as
she can possibly write. Suppose the right to Agatha’s labor is assigned to Sir
George. That means that Sir George can direct the way Agatha’s labor is to
be used: she is his slave. Sir George will, of course, be an enlightened slave-
holder, in the sense that he will not work Agatha so hard that the total value
of what she produces declines. But he will work her just short of that point.
Suppose that Agatha, if she had the right to her own labor, would work as
an interior designer, at which work she would make much less money but
find her life more satisfying. Or suppose that she would write many fewer
detective stories than she could, sacrificing the additional income to spend
time at her garden. At some point she would rather stop writing to enjoy
what she has made, rather than make marginally more money, but have no
time to enjoy anything. She may, perhaps, work somewhat more effectively
while she works if she is her own master— but she will probably work at a
less lucrative job, and almost certainly will work less.
If she tells the bank manager that she intends to design interiors, or to
work at her garden, she will not be able to borrow anywhere near the funds
necessary to buy the right to her own labor from Sir George. If she does not,
but leads her life that way anyway, she will soon be in default on debt ser-
vice. She can borrow enough money, even to make Sir George indifferent
about selling her the right to her labor, only by undertaking to lead a life as
distasteful to her as the life she would have led under Sir George. She will
have to perform almost exactly the labors that he, as a master of enlight-
ened self-interest, would prescribe. She will cease to be his slave only by
becoming the slave of the First National Bank (of Chicago, of course).
Indeed, her situation is even worse than that, because I have ignored the
interest the bank will take. (The rate may be high if others are at the same
time trying to find capital to buy back the right to their labor.) So her ability
to borrow enough to make Sir George indifferent will depend upon his
other investment opportunities, and (if he is confident about her abilities)
 Is Wealth a Value? 255

his risk aversion. Nor is it by any means plain that, if she could borrow
enough, she would. Sjhe gains very little actual control over the conduct of
her life, as we have seen, and she loses a considerable degree of security.
The main value of freedom is the value of choice and self-direction, and if
she starts her career a slave she will never be able to recapture more than a
token amount to these. We cannot be confident (to understate) that a thor-
ough analysis would justify the conclusion that Agatha either could or
would buy back the right to her labor. We therefore cannot claim that eco-
nomic analysis supports giving her that right in the first place.
Readers will no doubt think that I have gone mad some time ago. They
will think that the character of the arguments I have been making demeans
the case against the normative aspect of the economic analysis of law. Many
will think it more important to say that a theory that makes the moral value
of slavery depend on transaction costs is grotesque. They are right. But my
present point is not that wealth maximization, taken seriously, may lead to
grotesque results. It is the more limited point that this particular effort to
show that wealth maximization has strong instrumental value wholly fails.
Posner has another argument we should notice here. He gives some place
to a different instrumental claim: wealth maximization has value because a
society that seeks only social wealth maximization will encourage attractive
personal virtues, particularly the virtue of beneficence. This is not an un-
familiar argument. Defenders of capitalism often call attention to how the
“Protestant” virtues of industry and self-reliance thrive in a capitalist sys-
tem, but they do not give prominence to specifically altruistic virtues. It is
this feature of the claim that makes Posner’s account so paradoxically at-
tractive.
Posner’s argument is straightforward: in a society dedicated to wealth
maximization, people can improve their position only by benefiting others,
because when someone produces goods and services others buy, he must be
producing $pme benefit for them as well as for himself. The argument does
not specify the metric it assumes for testing whether a society bent on
wealth produces more beneficial-for-others activity than a society that en-
courages a more direct altruism. It is not easy to see which metric would be
appropriate. Even if wealth-produced-for-others is taken as the measure,
with no allowance for distribution, it is far from clear that more wealth will
be produced by people for other people, as distinct from themselves, under
wealth maximization that under a system of taxation and redistribution,
even though the latter produced less wealth altogether. Surely, welfare-for-
others is a better measure of moral achievement than simply wealth-for-
others, and, because of marginal utility, welfare-for-others is a standard that
includes distributional requirements. It is far from plain that wealth maxi-
mization will produce more total welfare-for-others activity than other,
more compromising, economic and political structures.
But that is an empirical question. We need not pursue it here, moreover,
256 The Economic View o f Law

because of a more fundamental flaw in Posner’s argument that wealth max-

imization is of instrumental value because it produces people who benefit
others. For the moral value of beneficial activity, considered in itself, con-
sists in the will or intentions of the actor. If he acts out of a desire to im-
prove the welfare of others, his act has inherent moral value even if he does
not benefit others. But it has no inherent moral value if he acts with the in-
tention of benefiting only himself. Posner makes plain that his production-
for-others claims have nothing to do with the other-regarding intentions of
actors in the economic process. He supposes, on the contrary, that they will
act to maximize benefit to themselves, benefiting others only through their
inability to absorb every last bit of the consumer surplus, as they would like
to do for themselves. The better someone is at personal wealth maximiza-
tion— the more he displays the skills and talents to be rewarded in the sys-
tem— the less his acts will benefit others, because the more of the surplus
will he be able to retain from each transaction or enterprise. Any benefit to
others comes from the invisible hand not good will. It cannot be the intrin-
sic value of wealth-producing acts that recommends wealth maximization.

I t i s , pe r h a ps , the consequences of these acts. Perhaps the individuals

seeking wealth only for themselves will produce a distribution that is just.
This suggestion, in its widest scope, supposes that a society pursuing wealth
maximization will achieve a closer approximation to ideals of distributive
justice, for some reason, than a society not single-mindedly pursuing that
goal. These ideals of distributive justice must be specified, or at least con-
ceived, independently of wealth maximization. It will not do to say that dis-
tributive justice is whatever state of affairs is produced by wealth
maximization. For then the claim that wealth maximization leads to distrib-
utive justice would be merely tautology.
So this new interpretation of the instrumental account must be com-
pleted by at least a rough specification of justice. It would be natural for an
economic analyst to choose one of the several accounts of justice already in
the traditions of political philosophy— highest total or average utility, for
example, or equality, or maximin over welfare or wealth, or some merito-
cratic theory. The theory selected must be a patterned rather than a histori-
cal theory, to use Robert Nozick’s useful distinction.13 Historical theories
argue that a distribution is just, whatever inequalities or other features it
displays, if it is reached in accordance with correct principles of justice in
acquisition and transfer. Patterned theories argue that a distribution is just
only if it conforms to some pattern that can be specified apart from the
history of how that distribution occurred. Wealth maximization specifies a
patterned rather than a historical test for the assignment of rights: the deci-
sion whether the doctor or the candymaker has the right each seeks is to be
 Is Wealth a Value? 257

made with a pattern in view— goods should be in the hands of those who
would pay most to have them. It is almost incoherent to propose that a pat-
terned distribution might be instrumental in achieving a historically con-
tingent distribution.
The defender of wealth must thus choose some patterned conception of
justice, like highest utility, equality, maximin, meritocracy, or desert.
Posner disclaims the first three of these specifically. Merit or desert theories
are more congenial to his spirit, so we will consider these first.
Meritocratic theories hold that justice consists in that distribution in
which people are rewarded in accordance with their merits. We now sup-
pose that wealth maximization might be said to be of strong instrumental
value, because (through an invisible hand or false-target mechanism) a so-
ciety whose laws seek only wealth maximization will produce the required
meritocratic distribution, or come closer to it than any alternative system.
But we must now distinguish between two conceptions of merit such an ar-
gument might employ. The first we might call an independent conception
of merit. It requires that we be able to state what counts as merit indepen-
dently of the wealth maximization process, so that it becomes an empirical
hypothesis that wealth maximization rewards merit so stated. But for any
list of independent merits that empirical hypothesis must fail, because
which abilities or traits will be rewarded in any particular community at
any particular time is a matter of technology, taste, and luck. Consider that
set of talents necessary consistently to hit a breaking pitch. If any list of in-
dependent merits does not include that set, then it will be false that in our
society wealth maximization rewards merits better than alternatives. Ted
Williams will be rewarded, in such a system, much more highly than almost
everyone else who ranks higher in the set of merits we do list.
If, however, we list that set of talents as merits, it will be false that
wealth maximization characteristically rewards merits. That set of talents
was not rewarded before baseball developed as it has, is not now rewarded
where baseball has not so developed, and will not be rewarded if baseball
declines and disappears. We can generalize: since which talents are
rewarded by the market is highly contingent on a variety of factors, the
pursuit of efficiency cannot be relied on to reward any particular set of
these fixed as independent merits over time. But neither can it be relied on
to disregard any particular set.
I shall call the second conception of merit the dependent conception. It
holds that merit is constituted by the set of talents that enables one to suc-
ceed in the market from time to time. Some of these talents are relatively
fixed, such as industry, shrewdness, and, perhaps, greed. Typically, although
not inevitably, one does better with industry or shrewdness than without it.
Other talents become merits only by virtue of transient tastes and luck; they
are merits for a time because they enable one to produce what others take
258 The Economic View of Law

to be benefits to themselves and are willing and able to buy. Under the de-
pendent conception of merit, it is true that a market economy geared to
wealth maximization will reward merits. It is too true, for under the depen-
dent conception the instrumental claim has collapsed into tautology.
At least for Posner, therefore, we cannot find any suitable independent
conception of justice in the literature of political philosophy. He makes a
wide claim for wealth maximization, but he has rejected all the conceptions
that do not make that wide claim either plainly false or trivial. What about
pluralistic conceptions of justice? I mean theories that disclaim any single
value, like utility or equality or merit, as making up all of justice in distri-
bution, but instead argue that a truly just distribution will achieve a sensible
mix of several of these values. The just distribution, on a pluralistic concep-
tion, will be one in which the average level of welfare is reasonably high, in
which there is not too much inequality, and in which what people have is at
least roughly related to how hard they have worked or how much they have
produced. It may not be possible to specify the exact mix of the different
components of the just society. But someone may claim to know it when he
sees it. Is it sensible to say that wealth maximization is instrumentally re-
lated, in the strong sense, to some such pluralistic conception of justice?
The danger is evident enough. The instrumental claim completed in this
way is in danger of becoming a tautology once again, unless the pluralistic
conception is stated clearly enough to allow that claim to be tested empiri-
cally. That is close to impossible. Let us suppose that single-minded wealth
maximization, in a particular society, would produce a certain precise car-
dinal level of average utility, a specific inequality factor (measured, for ex-
ample, in Gini coefficients), and a determinate correlation between merit,
somehow defined, and wealth. A critic now proposes a compromise with
wealth maximization— for example, by a piece of redistribution that lowers
the total wealth of the community. That compromise would produce
slightly less average utility, slightly less inequality, and a different correla-
tion between merit and wealth. Each of these factors, that is, becomes
somewhat, but not radically, different from the result under single-minded
wealth maximization. Now the partisan of wealth maximization on this in-
strumental argument must suppose that the original mix of these different
components of social value is better than the new mix. It is not enough for
him to suppose that the original mix is better than the maximand of any of
the three components: better than the society in which average utility is as
high as possible, or inequality as low as possible, or people are never
rewarded except in proportion to merit. He must also believe it better than
the different mixes of these three desiderata that would be achieved under
political and economic systems less uncompromising than his single-minded
wealth production.
His belief is implausible. It is highly indeterminate, ex ante, which car-
 Is Wealth a Value? 259

dinal level of average utility, which coefficient of inequality, and which

correlation of wealth, to merit (on any nontautologous definition of merit)
will be produced by a program of wealth maximization. It is also highly in-
determinate which mix of these putative desiderata would be achieved by
any discrete compromise with wealth maximization. It is, therefore, im-
plausible that a particular mix exists such that it is both independently pref-
erable, on moral grounds, to possible alternatives, and also antecedently
more likely to be secured by wealth maximization than by discrete com-
promises. My point is not that it is impossible antecedently to describe the
“best” mix of components, other than in the I-know-it-when-I-see-it fash-
ion— although that is a bad sign. But rather that, at the level of fine tuning
necessary to distinguish the results of wealth maximization from the results
of compromises, there simply is no one “best” mix antecedently more likely
to be produced by one rather than the other of these social techniques. The
pluralistic instrumental account is weaker than a straightforward hitching
of wealth maximization to a traditional theory— for example, utilitari-
anism— might be. In the latter case the goal that different instrumental
theories compete to maximize is at least specifiable.
There is an important and more general point here. Even patterned
theories of justice are likely to leave something to the contingencies of his-
tory. At a certain level of fine tuning, for example, even a strict egalitarian
will admit that the result of a trade between equals respects equality just
because it is a trade among equals, rather than because its results are those
specifically demanded by equality. I suspect that partisans of wealth maxi-
mization also believe that a particular distribution is just because it is the
distribution achieved by wealth-maximizing rules, and not vice versa.
Surely that suspicion is supported by the great bulk of writing exploring the
economic analysis of law. But of course that judgment takes us back to
wealth as a component of value. It cannot be supported by any instrumental
defense of \yealth maximization. It supposes, instead, that wealth maximi-
zation is a fair procedure whose results are just, as an egalitarian supposes
that a trade between equals is an inherently fair procedure. So a wealth
maximizer who holds that a distribution is just if it is the product of wealth-
maximizing rules cannot rely on any instrumental justification of at least
that aspect of his theory.

W e h a v e b e e n considering how the various forms of the instrumental claim

for wealth maximization might be completed by specifying an independent
conception of social value that wealth maximization promotes. I first set
aside the utilitarian conception of justice because Posner explicitly rejects
that conception. But Posner’s own suggestions— individual rights, individ-
ual virtue, and some impressionistic mix of different values— all fail, and al-
260 The Economic View o f Law

though he has been the most explicit and extreme wealth maximizer among
lawyers, his rejection of utilitarianism is not binding on the others. Does the
utilitarian tradition offer a way of completing the instrumental defense of
wealth?
I do not mean, in raising that question, to endorse utilitarianism in any of
its various forms. On the contrary, it seems to me that utilitarianism, as a
general theory of either value or justice, is false, and that its present unpop-
ularity is well-deserved. It is not, however, a theory that can be rejected
out-of-hand, by any argument as simple as the argument I used to dispose of
the theory that wealth is a value in itself. It has enjoyed the support of a
large number of sophisticated and sensitive philosophers. It is, therefore,
worth asking whether a thoroughgoing utilitarian might be led to support
wealth maximization on an instrumental basis.
Once again we must be sensitive to the different types of instrumental
theory. There are invisible-hand, ingredient, and false-target versions of the
instrumental thesis available, and also versions of wider and narrower
scope. The versions share, however, a common conceptual problem. Utili-
tarianism supposes that individual welfare levels are at least sometimes
comparable, so that total or average utility levels can be ordered over vari-
ous choices of social programs. Economists as a group have been skeptical
about interpersonal comparisons of utility. If utilitarianism is to be the
motor of wealth maximization, then wealth maximizers must forego that
skepticism and move even further from present orthodoxy in economics.
But when we admit generalizations about comparisons of welfare within
large communities— like the generalization that the marginal utility of
wealth declines— then any broad version of the utilitarian-instrumentalist
theory becomes immediately implausible. It is implausible to think that a
society that seeks wealth maximization single-mindedly will achieve more
total utility than a society that seeks wealth maximization but puts an upper
bound on the level of inequality it will tolerate in the name of social wealth.
So any plausible utilitarian-instrumentalist theory of wealth maximiza-
tion must be a reasonably narrow theory. Let us construct a sample narrow
theory tied to adjudication. This holds that a society whose judges decide
hard cases at common law by choosing the rule expected to maximize social
wealth will achieve more total utility in the long run than a society that
chooses another discrete program for deciding such cases, including a so-
ciety whose judges decide such cases by choosing the rule that can be ex-
pected to maximize total utility in the long run. This is a strong
instrumentalist theory; it defines a group of political decisions (hard com-
mon law cases) such that officials are required to decide all such cases to
maximize wealth, rather than to ask, in each case, whether maximizing
wealth in that case would promote utility. What sort of empirical evidence,
or set of correlative assumptions, would support that theory?
 Is Wealth a Value? 261

The most eligible assumption considers selective wealth maximization as

an ingredient rather than as a cause or false target of value. It supposes that
if judges decided such cases so as to increase the total wealth, other institu-
tions— perhaps legislatures— would then redistribute the increased total
wealth to improve average or total utility. That chain of events is, no doubt,
conceivable, once we accept that interpersonal comparisons of individual
utility make sense in principle. It is not, however, inevitable. The political
process might, for a variety of reasons, leave those who gain most from
wealth maximization with their gains intact. We should, therefore, ask
whether the utilitarianism-instrumental theory requires that legislatures
actually redistribute to improve total utility, or whether it is sufficient, to
support that theory, simply that they might do so.
Consider the following elaboration of the theory. Judges decide discrete
common law cases against the background of a given distribution of wealth
and legal rights. No decision a judge makes in a particular case will signifi-
cantly affect that distribution. The best a judge bent on improving total util-
ity can do is to improve the total supply of wealth. If the legislature finds
some way to redistribute the increased wealth so as to optimize utility, well
and good. If not, nothing has been lost. It is better to provide the legislature
with an opportunity to improve utility, even if the opportunity may not be
taken, than to do nothing.
Is this a good defense of our narrow theory? It rests on a large assump-
tion: that there is nothing that judges can ever do directly to advance utility
more than they can do simply by maximizing wealth, even when they know
that the legislature will do nothing further to advance that goal itself. It as-
sumes that judges would promote utility less overall, even in those circum-
stances, if they sometimes asked whether a less single-minded, more
discriminating approach would improve utility in particular cases. It rests
on the assumption that wealth maximization is a good false target for util-
ity even when it is not a useful ingredient of utility. We may test that
assumption Tn this way. Suppose someone suggests the following alterna-
tive program for adjudication. Judges should reach that decision, in hard
cases at common law, that will promote utility better than any alternative
decision. In some cases, perhaps most, that will be the wealth-maximizing
decision and in some not. Everything depends on circumstances, and it is
impossible to say in advance how often this theory will recommend non-
wealth-maximizing decisions.
That is (in the sense defined) a weak instrumental theory of wealth maxi-
mization. Two questions arise. Will the weak theory ever recommend a ju-
dicial decision the narrow strong theory would not? Will a society whose
judges follow the weak theory produce more utility in the long run than a
society that follows the strong one? The answer to the first of these ques-
tions will depend on a variety of issues, but is almost certainly, yes. Pater-
262 The Economic View o f Law

nalism will provide occasions when the utility-maximizing rule differs from
the wealth-maximizing rule. Suppose, for example, that the community will
pay more for candy than for medical care lost through the noise of a candy
machine, but the candy will be bad for its health and therefore its long-term
utility. Future generations provide other occasions: once the utility of fu-
ture generations is taken into account, even common law decisions— like
those affecting the environment— may injure utility if they promote wealth
in its present distribution. Quite apart from these factors, some common
law decisions are potentially redistributive. Suppose a decision might either
protect the workers of an ailing and, possibly, noncompetitive industry or
hasten their unemployment by structuring rights in favor of a develop-
ing new industry? The wealth-maximizing decision might be the latter; the
utility-promoting decision nevertheless the former.
If there are many occasions on which the two theories— weak and
strong— would recommend different decisions, the answer to the second
question is probably, no. It is true that false targets are sometimes good tar-
gets: we sometimes gain more by aiming slightly away from what we want,
as a man bent on pleasure would do well not to aim directly at it. But that is
not always or even usually so, and there seems, a priori, no more reason why
it should hold in the case of courts than in the case of legislatures. If it is
sometimes true that a legislature should choose a decision that does not
maximize wealth, because it will nevertheless improve utility, there seems
no reason why a court should not do so as well. The occasions on which a
court has that choice are, perhaps, fewer, but that is plainly a different
matter.
So the utilitarian-instrumental theory does seem to depend on some
judgment that the legislature will act in cooperation with courts to redis-
tribute, so as to produce more utility from the wealth the court provides.
But if that is so, then the theory is seriously incomplete, because, so far as I
know, that case has never been made. Nor is it immediately plausible. On
the contrary, if the familiar assumption is right, that optimal utility would
require much more equality of wealth than now exists in our country, the
hypothesis that the legislatures, federal and state, have been busy redistrib-
uting in search of utility seems embarrassingly disconfirmed.
Even if that hypothesis were sound, much more would be needed to de-
fend judicial wealth maximization in this way. We should still have to show
why, when more utility could be produced by a decision aiming directly at
utility, the court should aim instead at wealth. The hypothesis, that the leg-
islature will concern itself with utility, is not in itself a satisfactory answer.
Would the gains in utility not be provided sooner and more securely in one
step rather than two? There seems no reason not to prefer a weak instru-
mental theory: courts should decide to maximize utility, recognizing that
the existence of legislatures ready to redistribute might mean that on some
 Is Wealth a Value? 263

occasions wealth improvements might be the best means of improving util-

ity in the long run. If any strong theory is preferred to that weak theory, it
must, once again, rest on the (unsupported) false-target theory.
I have considered, in this part of the essay, whether a strong instrumental
theory can defend wealth maximization, taken as the single-minded goal for
at least a discrete part of adjudication, on the assumption that total utility is
a value in itself. I argue that the hypothesis that it can seems weak, and is
far from demonstrated. The same arguments apply, I think, against any
strong instrumental claim for wealth maximization that takes maximin (on
either the wealth or the utility space) rather than total utility to be a social
value in itself. Once again, the question is raised why a weak theory, which
encourages judges to seek maximin solutions directly, taking due account of
the potential instrumental role of wealth maximization, would not be supe-
rior. No answer to that question has been provided, and it is not clear that
there is a good one.
I should close this section, however, by noticing what I hope has been
apparent in the discussion so far. The instrumental claims for wealth maxi-
mization are more plausible if they are harnessed to one of the nonmerito-
cratic-patterned theories of justice, such as utility or maximin, than to
anything else. They cannot then be ruled out conceptually as, for example,
Posner’s instrumental claims can be. But they are still— certainly in the
present state of play— claims with almost no foundation.

E c o n o m i c a n a l y s i s of law is a descriptive and a normative theory. Does

the failure of the normative limb impair the descriptive limb? The latter
offers an explanation of one aspect of human behavior, namely, the deci-
sions of common law judges in the cases economic analysis purports to ex-
plain. There are several modes (or, as some would say, levels) of explanation
of human behavior. Some of these are nonmotivational. These include ge-
netic or chemical or neurological accounts of either reflex or reflective be-
havior. The motivational modes of explanation may also be of different
forms. The most straightforward is explanation from the agent’s point of
view, an explanation that cites the agent’s goals or intentions and his belief
about appropriate means. But there are more complex forms of motiva-
tional explanation. Invisible hand explanations, for example, suppose that
people act out of certain motives, and explain why, that being so, they col-
lectively achieve something different from what they aim at individually.
One class of Freudian explanations also assumes that people act out of mo-
tives, but holds that these motives are unconscious. These Freudian expla-
nations are, nevertheless, motivational because their explanatory power
hinges on the claim that people whose behavior is so explained are acting in
a way best expressed by analogy to the behavior of people who hold such
264 The Economic View o f Law

motives consciously. The theory is therefore dependent on an understand-

ing of that straightforward motivational claim.
The argument of economic analysis, that judges decide hard cases so as to
maximize social wealth, is not a genetic, chemical, neurological, or any
other form of nonmotivational explanation. Nor is it an invisible hand expla-
nation. It is true that something like an invisible hand explanation of why
common law decisions promote social wealth has been offered,14 but this is
not part of the claims of Posner, Calabresi, or other proponents of economic
analysis. To my knowledge, economic analysis has never been presented as
a Freudian analysis. But even if it had, that analysis would presuppose the
sense of a straightforward claim. So economic analysis, in its descriptive
limb, seems to rest on the sense and the truth of a straightforward motiva-
tional claim, which is that judges decide cases with the intention of maxi-
mizing social wealth.
But my arguments against the normative limb of economic analysis also
call any such motivational claim into question. I did not argue that maxi-
mizing social wealth is only one among a number of plausible social goals,
or is a mean, unattractive, or unpopular social goal. I argued that it makes
no sense as a social goal, even as one among others. It is preposterous to
suppose that social wealth is a component of social value, and implausible
that social wealth is strongly instrumental toward a social goal because it
promotes utility or some other component of social value better than would
a weak instrumental theory. It is, therefore, bizarre to assign judges the mo-
tive either of maximizing social wealth for its own sake or pursuing social
wealth as a false target for some other value. But a straightforward motiva-
tional explanation makes no sense unless it makes sense to attribute the mo-
tive in question to the agents whose behavior is being explained.
It follows that the descriptive claims of economic analysis, as they have
so far been presented, are radically incomplete. If they are to have descrip-
tive power, they must be recast. They might be recast, for example, in some
way appropriate to a weak instrumental claim. The arguments must then
become more discriminating. They must pick out particular classes of judi-
cial decisions and explain why it was plausible for judges to suppose that a
rule improving social wealth was likely, for that reason, to advance some
independent social goal these judges valued— utility, maximin, the relief of
poverty, the economic power of the country in foreign affairs, or some other
goal. That becomes a claim of great complexity, for it involves not only a
detailed causal account but detailed intellectual history or sociology. Did
judges who developed the fault system in negligence or the system of strict
liability suppose that their decisions would advance average total utility?
Were these judges uniformly utilitarians, who would therefore count that
an advantage? Does this explanation hold good only for a certain group of
cases at a particular point in the development of the common law? Is it
 Is Wealth a Value? 265

plausible to suppose that judges throughout some extended period held the
same theory of social value? Is it plausible to suppose, for example, that
they were utilitarians indifferently before, during, and after the academic
popularity of that theory of social justice? That only scratches the surface of
the kind of account that would be needed to give a weak instrumental ex-
planation of judicial behavior along wealth maximization lines, but it is
enough perhaps to suggest how far short the present literature falls. It has
not achieved the beginning of a beginning.
It may now be objected, however, that I am asking for far too much, and
unfairly discounting what has been done already. Suppose that the eco-
nomic analysts have established an important correlation between the deci-
sions that common law judges have reached in some particular area— say
nuisance or negligence or contract damages— and the decisions that would
have been taken by judges explicitly seeking to maximize social wealth.
Suppose that, although not every decision actually made is the decision
such a judge would reach, the great majority are. (I know this putative cor-
relation is contested, and I assume it in this section arguendo.) It seems silly,
not to say churlish, to turn our backs on all this information. We may hold
the following attitude. No doubt it would be better still if an intellectual
historical account could explain why actual judges acted in this way, either
by showing that they took wealth maximization itself to be a component of
value, or because they held a strong instrumental theory of wealth maximi-
zation, or a weak instrumental theory that had the consequences discov-
ered. But the correlation, in and of itself, advances our understanding of the
legal process to an important degree.
I think this attitude is wrong. It is wrong because a correlation of this sort
has no explanatory power unless it is backed by some motivational hypoth-
esis that makes independent sense. Suppose the following exercise. Let us
construct a binary alphabetical priority sequence for all cases ever decided
by the highest court in Illinois. (We take 1 if the winning party’s name is
alphabetically prior to the loser’s; 0 otherwise. Forget complications or
ties.) Call the sequence Arthur. We would not say that Arthur explains the
judicial decisions in these cases, although Arthur is in fact a perfect correla-
tion. Arthur has indefinitely many projections into the future. Suppose each
academic lawyer in the United States were to project Arthur to a further
100 places at random. We would then have a very large variety of further
sequences (Arthur Posner, Arthur Michelman, Arturo Calabresi, and so on),
one of which would predict the results of the next, say, 100 decisions on the
Illinois court better than any other, and, quite likely, very well indeed. But
we would not say that, for example, Arthur Michelman had great predictive
power or was a better theory of judicial decision making in Illinois on that
account.
The point is both evident and important. Our standards for the explana-
266 The Economic View of Law

tion of human behavior require, in order for some account even to be a

candidate for an explanation, that it bring to bear either a biological or a
motivational account. If a correlation, however secure, cannot promise
even the prospect of such a connection— if these connections cannot sensi-
bly be taken even as mysteries waiting to be solved— then it becomes coin-
cidence only. The claims for astrological and other occult explanations of
behavior are problematical in this way. It strikes many people that both a
motivational and a biological account are excluded by positive conclusions
of physics that are beyond reexamination; but it strikes others that Hamlet’s
warning to Horatio is sound and pertinent.
We have three choices. We may disregard the putative correlation be-
tween actual and wealth-maximizing decisions as coincidental and attempt
to construct theories of adjudication that ignore it. That seems wasteful and
perverse, for the correlation, if it exists, is different from the correlation be-
tween Arthur and the cases from which Arthur was constructed in one im-
portant respect. In the case of Arthur, the method of construction
guarantees that the correlation is coincidental rather than explanatory. In
the case of economic analysis, coincidence is one hypothesis only.
Second, we may pursue the enterprise I suggested earlier in this section.
We may try to construct a weak instrumental theory of wealth maximiza-
tion showing why, in just the areas of law where the correlation holds, the
weak instrumental theory, harnessed to some conventional idea of social
value like utility, would recommend the wealth-maximizing strategy as a
good means, and why it is plausible that judges realized this, in at least a
rough and inarticulate way. That enterprise would carry economic analysis
into layers of detail of both political theory and intellectual history it has
not yet even begun to reach. But the enterprise cannot be dismissed in ad-
vance.
There is a third choice. We may try to embed the correlation in a radi-
cally different sort of analysis and explanation. We may try to show that the
decisions that seem to maximize wealth are required, not as instrumental
decisions seeking to produce a certain state of affairs, of social wealth, util-
ity, or any other goal of policy, but rather as decisions of principle enforcing
a plausible conception of fairness. We might aim, that is, at an explanation
of principle, instead of an explanation of policy. I have, on various occa-
sions, tried to show why an account of judicial decisions on grounds of prin-
ciple should always be preferred to one on grounds of policy, for normative
and positive reasons. I have also illustrated a strategy for a principled ac-
count of judicial decisions that look to consequences, including third-party
consequences.15 This strategy of principle seems to me much more promis-
ing than the weak instrumental program of policy just described.16 But I
have not yet provided any compelling reasons why you should join me in
that confidence.