Roy Cordato’s “Austrian” Critique of Coase on Social Cost*
July 18, 2000
Abstract
In his book on Austrian welfare economics, Roy Cordato presents a scathing criticism of
the Coasean approach to making judgments in tort cases. His criticism claims that the Coasean
approach is both unrealistic and normative. Moreover, it would introduce uncertainty, thereby
reducing individuals’ freedom in pursuing their goals through the exchange process. To avoid all
of these problems or at least to better deal with them, the economist who is interested in giving
advice to judges ought to advocate a rule of strict liability regarding invasions of real (i.e.,
material) property. He claims that this rule is derived from Austrian economics.
This paper, which takes a “different” Austrian perspective, makes several arguments
against Cordato. First, it argues that his recommended rule would not achieve the goal that he sets
for it. It would increase the freedom of some people while reducing the freedom of others in
pursuing their goals. Second, it argues that his criticism of Coase is misplaced. Contrary to
Cordato’s claim, Coase has not advocated day-to-day interventions in the market system based on
benefit-cost calculations that assume general competitive equilibrium. On the contrary, like the
Austrians, he has emphasized the difficult knowledge and calculation problems associated with
this. Also, he has pointed out the preponderance of empirical evidence against such interventions.
Moreover, Coase’s recommendations do not concern day-to-day interventions at all. Rather, he
has been concerned with the institutional structure, most especially with the initial delimitation of
rights. This concern, which implicitly takes account of uncertainty, immunizes Coase’s analysis
from Cordato’s uncertainty critique.
In a book that claims to present a modern Austrian perspective on welfare economics and
externalities, Roy Cordato introduces a welfare criterion that is “divorced from perfect
competition or any notion of general equilibrium.” He believes that such a criterion will provide a
“sound theoretical basis for further empirical and public policy analysis in the area of externalities”
(Cordato 1992: 1). Other neo-Austrian economists have proposed welfare criteria.
The
distinguishing feature of Cordato’s proposal is his scathing criticism of Ronald Coase’s paper on
social cost (Coase 1988a [1960]). He claims that the Coase approach to public policy advocates,
or at least requires, micro-management of economic affairs. This requirement disregards the
established Austrian view that no one could possess the knowledge needed to make such
decisions. Since micro-management is impossible, he writes, anyone who tries to apply the
criterion must make an ethical judgment. Of course, Cordato’s welfare criterion also entails an
ethical judgment. However, it does not require micro-management.
This paper makes two arguments. The first is that Cordato’s own analysis is inconsistent.
The ultimate goal of his welfare criterion is to enable individuals to more efficiently pursue their
goals through the exchange process (Cordato 1992: 64). Yet the particular judgments he
recommends in tort cases are unlikely to achieve this goal. Although they would facilitate the
pursuit of goals by one or more persons in the exchange process, they would also hinder the
pursuit of goals by others. The source of the problem lies with the limited relevance of Cordato's
material conception of property and factors of production, a conception that Coase himself sought
to change among economists.
The second argument of the paper is that Cordato's assumption that Coase’s goal was to
provide a criterion to help make day-to-day policy and judicial decisions (i.e., to micro-manage) is
incorrect in two ways. First, Coase’s paper was, for the most part, intended as a criticism of Pigou
and the Pigovian approach. Coase wanted to evaluate Pigovian policy proposals on the basis of
whether they are logically consistent and relevant, given the stated goals of Pigou himself. Such
an aim is broadly consistent with Ludwig von Mises’s prescription for making economics value-free.
Thus, Cordato’s claim that Coase’s effort implies normativism is mostly wrong. This is
discussed in part two.
Second, Coase’s focus on judicial decisions rather than legislative decisions demonstrates
his interest in institutional structure, or design, and not in day-to-day micro-management. Coase
was concerned specifically with judicial precedents. This is important because Cordato criticizes
Coasean analysis on the grounds that, relative to tort decisions based on physical property rights,
Coasean recommendations could introduce uncertainty (Cordato 1992: 104-5). To avoid this,
Cordato argues, tort decisions should adopt the strict liability rule based on property rights. Part
three compares the strict liability rule with the rule of negligence and shows how the latter is more
appropriate for a “dynamic,” uncertain world in which rights to control actions that were not
previously recognized as factors of production are continually created.
1. Cordato’s Welfare Criterion and the Institutional Means of Achieving It
Cordato’s Welfare Criterion
The basis of Cordato's welfare criterion is the judgment that individuals should be able to
pursue goals, even if they make errors. He writes that "[t]he task of the economist, when
considering normative questions, is to identify those institutions that best facilitate [the process of
trial and error]." "[T]he overriding purpose...is to identify, generally speaking, the framework
within which individuals, as social beings, are able to most efficiently [i.e., through the trial and
error process] pursue their goals" (ibid.: 65). Note that he uses the term "pursue" and not
"achieve" (ibid.: 68-69). Although he does not point this out, it is evident that an individual could
totally fail to achieve any goals and feel so miserable that he would commit suicide. He may also
prefer a system in which freedom to pursue goals is hampered in some way. Nevertheless, if
institutions enabled him to pursue his goals through the trial and error process, the institutions
would be judged as good.
He labels his criterion for judging institutions “catallactic efficiency.”
With this welfare criterion in mind, he goes on to describe two characteristics of what he
calls the ideal institutional setting (IIS) (Cordato 1992: 64-8). The first characteristic is private
property. The second is freedom to exchange. Such a setting will "best facilitate the use and
discovery of information, the appropriateness and relevance of which can only be known by those
who need to discover and use it." And it will "allow individuals to gather the necessary physical
resources" to carry out their individual plans (ibid: 63).
Regarding how such an ideal should be applied, Cordato refers to certain "institutional
inefficiencies...that are generated by deviations from the 'ideal' legal framework" (ibid.: 69). He
writes that such inefficiencies "can arise because people are prevented in some way from fully
utilizing their property or because rights to property are not clearly delineated” (ibid.). Cordato
admits that the IIS is difficult to fully specify. However, in light of his attack on an alternative
framework, it would seem that he should provide at least some guidance on key issues. First, he
does not consider whether a judge should attempt to correct “institutional deviations” from the
"ideal legal framework." For example, should a judge take away a monopoly franchise that had
been granted by a previous regime? Second, there is a question of what it means to say that rights
to property are not clearly delineated. He deals with this issue in one example: that of the
railroader’s sparks and the farmer’s land. He writes that when titles are not clearly delineated, the
judge should try to delineate them on the basis of “land surveys and the nature of deeds or
contractual arrangements such as protective covenants.” When titles are clearly delineated, the
judge should just enforce them. “If the farmer had title to the land that the crops were growing
on, there is no question that his right to continue to grow them should be upheld” (ibid.: 100).
The problem is either that he leaves the meaning of “clearly delineated” unclear or that the
meaning he uses is only partly relevant. What he apparently means is that existing material
property titles imply, unless bargained away, a right to receive compensation for all decreases in
value for which a “but for the presence of another party” test applies (ibid.: 102). However, it
may be that no one previously conceived of the kind of damage for which a judge is now being
asked to hold the other party liable. If so, a systematic ruling in favor of rights to material
property would provide a premium on the ownership of such property and a corresponding
penalty on the ownership of non-material factors of production. The importance of this point will
become more evident in the discussion that follows.
Railroad Sparks and Crop Fires
We now proceed to analyze the sparks and fires case in greater detail. Cordato writes the
following:
As opposed to Coasean analysis, where property rights are the most important variable, the Austrian
school’s approach to all externality related issues has consistently been that clearly delineated property
titles and rights must be taken as given. The (Austrian) notion of strict liability that is derived from that
view of property rights casts up a “but for” test of causality that subverts these conceptual ambiguities.
Causality is established, at least prima facie, if, but for the fact that the plaintiff (P) invaded the property
rights of the defendant (D), the harm would not have occurred. In this case “invade the property rights”
means to make use of property that D had title to, without D’s permission. In the case of the railroad’s
sparks and the farmer’s crops, again assuming that the farmer had title to the land on which he was
growing the crops, the approach offered here would lead to the conclusion that the railroad caused the
damage that was experienced by the farmer. But for the fact that the railroad’s sparks crossed over onto
the farmer’s land, the crops would not have been destroyed (ibid.: 102).
To see why this reasoning is deficient, suppose that an unprecedented and non-anticipatable
drought makes a farmer’s crops vulnerable to passing train sparks. Assume that prior to this time,
the sparks and crops had existed side-by-side, with no external effects. In the quoted statement
and surrounding text, Cordato implicitly asserts that in this kind of example, the judge should
regard the railroader’s action an “invasion of property.” The IIS judge should hold the railroader
liable for his invasion with one exception. This is the possibility that the farmer would have come
to the nuisance, which is not relevant here (ibid.: 103-4). Such a judgment would, in Cordato’s
view, facilitate the achievement of individual goals through the exchange process.
However, if one focuses on rights to control actions and not on rights to material
property, this conclusion does not follow. The railroader previously had the right to operate his
train near the farm. But the judge who follows the IIS rules is required to disregard this right. He
should only pay attention to rights to value embodied in the ownership of physical property, not
in its use. There is no doubt that the judgment implied by IIS rules would facilitate the farmer’s
pursuit of his goals through the exchange process. The farmer could proceed to borrow money
and to make advance contracts to deliver his crops with certainty that there would be no
uncompensated damage due to others’ actions in the “but for” sense. In addition, others who are
part of the network of exchange involving the farmer’s crops would be more able to pursue their
goals since they would have opportunities that would otherwise be unavailable. However, it is just
as evident that following IIS rules would reduce the railroader’s pursuit of her goals through the
exchange process. She would suddenly face the prospect that some change in environmental
conditions, or even market conditions, would thrust her into the role of a property invader. She
would be certain that unanticipatable changes in the environment or market of the type in question
would result in a liability judgment against her. She, her trading partners and others would be less
able to pursue their goals. The increased market value stability of the farmer’s land would seem to
be exactly offset by the reduced market value stability of the railroader’s right to run his trains in
the usual way.
Rights Associated with Physical Property vs. Rights to Control Actions: Faulty Concept of
a Factor of Production
The problem with Cordato’s analysis is that he uses what Coase called a faulty concept of
a factor of production (Coase 1988a: 155). In the example of the sparks and crop fires, an
unpredictable natural event is responsible for the creation of a new and valuable factor of
production. This factor of production is not a piece of physical property but a right to control the
railroader's spark-making actions. There is no reason to suppose that an institutional structure that
always assigns such rights to the material property owner, as opposed to someone else, would
promote catallactic efficiency.
More fundamentally, Cordato’s error appears due to his neglect of the issue of assigning
initial entitlements to non-material factors of production -- i.e., to rights to control actions. In the
continually changing entrepreneurial society, new unanticipatable factors of production are
continually created. Suppose that a new factor consists of property. Given (1) that the property
had been produced without employing coercion or fraud and (2) that the new property or its
production had no negative, non-pecuniary external effects on others; there would appear to be
little dispute that the creator-producer should possess the initial entitlement under the welfare
criterion that Cordato suggests. The same is true for produced non-material factors (i.e., rights to
control actions). However, when a new factor is a consequence of unanticipatable changes in
nature or when it or its production has negative, non-pecuniary external effects; even judges who
subscribe to the principle of catallactic efficiency would have to decide on the initial entitlement.
There is no more reason to suppose that assigning such a right automatically to a material
property owner would facilitate the pursuit of individual goals in a general sense than assigning
the right automatically to someone who does not own material property. It is to Coase’s credit
that he supplied both a definition of a factor of production and an analysis based on transactions
costs that would enable one to construct criteria to make such initial entitlement decisions.
Cordato gives no significant guidance on the question of how to assign such rights.
Coase’s approach assumes that when a legal process (lawmaker or judge) is faced with the
problem of deciding who should possess the initial entitlement, it should take into account the fact
that the entitlement may later be traded away. Given this fact, one of its aims should be to try to
assure that the entitlement is given to the individual who would be most likely to end up with it
after trading if transactions costs were zero. Coase does not actually advocate the use of judges to
determine initial rights based on transactions costs. However, this position is a natural implication
of two of his remarks:
The definition of rights that individuals and organizations are deemed to possess is important because,
by setting the starting point, it determines what transactions have to be carried out to achieve any other
constellation of rights, and therefore the costs of so doing (Coase 1988b: 673).
I concluded in “The Problem of Social Cost” that the value of production would be maximised if rights
were deemed to be possessed by those to whom they were most valuable, thus eliminating the need for
transactions (ibid.).
On the other hand, it should not be surmised from this that Coase advocated taking away the
rights of some people and giving them to others who value them higher. The topic under
discussion here is that of establishing initial entitlements when there is no other clear basis for
assigning them. Nothing the author has read would indicate that Coase advocates judges
reassigning existing entitlements unless the current owner had acquired the entitlements illegally.
Indeed, in a 1974 paper, he directs the reader’s attention to one aspect of the benefit-cost analysis
that is often disregarded, namely, that although there are indeed cases where government
regulation has improved the situation (accordingly to recognized standards of the meaning of
improvement), a correct evaluation of any proposed regulation requires the inclusion of all costs,
including the costs of carrying out the regulation. And these include the “costs” that we might
associate with the waste due to the self-interested actions of individuals in the political and
regulatory arenas (Coase 1974: 184-5). Also see Coase 1988a: p. 170-3.
In the Austrian Materialist “Tradition” Contra Menger
The focus on material factors of production, which is represented by Cordato’s
preoccupation with rights to material goods, is distinctly Bohm Bawerkian. Whereas Menger’s
definition of a good included such things as claims to future goods and money, goodwill, patent
rights, and franchise rights; Bohm Bawerk sought to purge these items from the definition of
computable wealth.
However, if one recognizes that the right to control actions and not the right
to control material factors of production is the more fundamental concept of a factor of
production in economic analysis, then it naturally follows that in order to implement any welfare
criterion, one must make judgments not about (or not exclusively about) who owns or should
own material property but about who owns or should own rights to control actions.
The point that Coase recognized and that had been neglected up to that time is that for the
purpose of logically analyzing judgments in tort cases, we ought to replace the vague notion of
“rights to property” with the notion of rights to control actions. “We may speak of a person
owning land and using it as a factor of production but what the land-owner in fact possesses is the
right to carry out a circumscribed list of actions” (Coase 1988a: 155). He goes on to point out
that this would be true under any system of law. It is suggested here that Coase’s approach to the
externality problem was more in the Austrian tradition of Menger than Cordato’s.
2. Coase’s Paper: Normative or Value-Free?
Referring to Coase’s social cost paper, Cordato writes that Coase "assumes that all
observed market prices are competitive equilibrium prices" (Cordato 1992: 93). He goes on to
claim that because these are not the prices that exist in a real market economy, the Coasean
analysis is impractical. When judges are faced with the problem of applying principles that Coase
derives on the basis of the competitive equilibrium assumption, they find their task impossible. "If
we place the judge within the context of a dynamic market process, where prices are not perfectly
competitive," Cordato writes, "his problems become insurmountable...[A judge] would have to
know all the opportunity costs associated with the use of the relevant resources for the point in
time that the market is being observed" (ibid.: 97).
In assessing Cordato’s evaluation, it is useful to distinguish two levels at which the
knowledge problem is manifest. The first is the level of individual action within the “rules of the
game” -- i.e., the knowledge of benefits and costs possessed by market participants. The second is
at the level of establishing the rules of the game -- i.e., knowledge of the benefits and costs
possessed by the judge or policy maker charged with making judgments about who should own
initial rights to control actions that have external effects. We discuss each in turn.
Knowledge at the Level of Individual Action
In presenting his farmer-rancher, railroader-farmer, and other examples, Coase assumed
for simplicity that the subjects of his study know the benefits of carrying out all of their possible
actions. He also assumed that all prices are competitive equilibrium prices, thereby removing any
possibility of market-calculation error on the part of the individuals. The principal theoretical
point that he wanted to make in these examples was that in a zero-transactions-cost world, the
choices of the parties were insensitive to the identity of the holder of the initial entitlements.
Under such conditions, actors’ post-exchange outcomes of choices would be the same regardless
of the initial distribution of the entitlements to control the performance of the relevant actions
(Coase 1988a:100-1). He then went on to discuss the case of positive transactions costs, in which
the post-exchange outcomes of choices might not be the same. He continued to assume that
individuals know the benefits of carrying out their actions and that there is a competitive
equilibrium. Thus, he assumed that there is no uncertainty by individuals about the outcomes of
their actions even though they face positive transactions costs (costs of discovering the relevant
prices and negotiation costs).
The point to be made here is that to isolate transactions costs from uncertainty in two-person exchange is a legitimate simplifying procedure, given that one’s goal is to elucidate the
influence of transactions costs. The theoretical point about the influence of transactions costs on
exchange outcomes in no way depends on assumptions made about uncertainty. If Coase had
assumed uncertainty, then he would have had to add tedious discussions based on various
assumptions about the nature of the knowledge differences among the individuals. And if he had
assumed that knowledge might be wrong, he would have had to add a tedious discussion about
the possibility that from the viewpoint of superior knowledge, the individuals might be better off if
they were not left to make their own decisions. Since his insight about transactions costs does not
depend on relaxing such assumptions, he did not do so. It is true that in everyday decisions about
exchange, transactions costs and uncertainty are inseparable; if for no other reason than that there
is uncertainty about transactions costs. However, in isolating the effects of transactions costs, he
abstracted from this connection. We conclude that Cordato’s criticism is irrelevant insofar as it is
aimed at Coase’s assumption about knowledge at the level of individual action.
Knowledge at the Level of Establishing Initial Rights to Control Actions
Cordato objects to Coase’s use of the term "social productivity" and to the assumption of
competitive equilibrium that he believes is necessary to make this term meaningful.
Coase’s use
of this term, Cordato argues, marks him as a neoclassical equilibrium theorist. Such a theorist
assumes that the policy analyst possesses sufficient information to determine equilibrium
outcomes. Cordato points out that this assumption is clearly at odds with reality because the real
market economy is "dynamic." Prices are always in disequilibrium (Cordato 1992: 97). He argues
that
in a truly disequilibrium world where the end state is a continuously moving target, it does not make
sense for legal rulings to attempt to insure any particular market outcome or distribution of resources.
The allocatively "efficient" outcome, if it can be meaningfully identified at all, is likely to be different at
each successive point, implying that it would only be by accident if any rule even tended toward
"efficiency" by the time it was actually implemented (ibid.: 98).
Coase himself did not make the error that Cordato attributes to him. True, in arguing
against the standard (Pigouvian) procedure of comparing private products with social products,
Coase writes that “[w]hen an economist is comparing alternative social arrangements, the proper
procedure is to compare the total social product yielded by these different arrangements” (Coase
1988a: 142). But he goes on:
The Pigouvian analysis shows us that it is possible to conceive of better worlds than the one in which
we live. But the problem is to devise practical arrangements which will correct defects in one part of the
system without causing more serious harm in other parts (ibid.).
Is it possible to devise such arrangements? In discussing the Pigouvian tax on the act of polluting,
Coase emphasizes that one ought to consider “the fall in the value of production due to the
pollution (in its widest sense). But to do so would require a detailed knowledge of individual
preferences and I am unable to imagine how the data needed for such a taxation system could be
assembled” (ibid.: 152, 182). The first item on his list of difficulties in implementing such a system
is “the problem of calculation” (ibid.: 152).
These remarks must be accounted for when contemplating the statement in the conclusion
to Coase's paper that “[i]t would seem desirable to use [the opportunity cost] approach when
dealing with questions of economic policy and to compare the total product yielded by alternative
social arrangements” (ibid.: 154). If one reads Coase’s paper carefully, it is evident that he
recognizes that there are difficult knowledge problems of the Austrian type associated with
making a judgment about which social arrangement is best.
It is true that he does not go as far as
some of the libertarian-minded Austrian economists to claim that no policy decision is justifiable
on social welfare grounds. It is also true that while he recognizes the difficulty of obtaining
information about preferences (ends), he does not mention information about the methods of
satisfying the preferences (means). However -- and this is the main point -- he does not make the
claim that Cordato attributes to him.
Mises vs. Ethics
I have elsewhere argued that neo-Austrian efforts like Cordato’s should be placed in
opposition to Ludwig von Mises’s claim that economics can be value-free.
In Mises’s view as an
economist, the derived aim of an economics of public policy analysis is to construct a logical
framework for evaluating the policy arguments of others. Using that framework, the economist
asks whether the proposed policy is likely to accomplish the goals that the proposer sets for it. So
long as economists confine themselves to this activity, they can remain value-free in a special
sense. They can (1) employ the universal rules of logic to make sure that the policy predictions
follow logically from the assumptions and (2) make judgments of relevance in an effort to
determine whether the predictions are consistent with the aims of the policy proposer. Mises, for
the most part, successfully maintained this value-free stance in his magnum opus Human Action
(1966). And, upon closer examination, Coase’s paper seems remarkably consistent with this view.
Coase did not make specific policy prescriptions; he assessed the prescriptions of others, based on
logic and relevance. He tried to show (1) that some of A.C. Pigou’s policy conclusions would
lead to consequences that are different from what he maintained and (2) that others were based on
assumptions that were irrelevant to the achievement of the policy goal Pigou proposed.
It is true
that Coase used the term “social productivity.” But he used it heuristically for the purpose of
dealing with the British tradition he was criticizing. Although this tradition might be properly
regarded as irrelevant or normative, it would be wrong on the basis of his “social cost” paper to
brand Coase a neoclassical welfare economist of the benefit-cost genre.
There is one characteristic of Coase’s reasoning that is not compatible with that of Mises,
however. Coase’s skepticism, as an economist, concerning the desirability of government
intervention stems much more from either his own historical research or from his reading of the
historical research of others (Coase 1996: 108). Mises’s skepticism, as an economist, stems
mostly from logic.
Before moving on to the next issue, it is necessary to deal with a puzzling claim by
Cordato that appears to contradict the argument of the previous paragraph. Without discussion,
Cordato claims that his notion of an institutional setting that “allows the greatest possible latitude
toward individual goal-seeking” is “ultimately derived” from Mises’s concept of value freedom
(Cordato 1992: 88 and 90n). He supports this claim with references. Unfortunately, most of his
citations are not from Mises’s mature economics but from his much earlier book on liberalism
(ibid.: 17). Mises’s book on liberalism was not intended to be value free. It was intended to
present the distinct value position of what Mises perceived as the liberal movement. Indeed, Mises
begins the section to which Cordato refers with the statement “As the liberal sees it...” (Mises
1978: 52). The two pages that Cordato cites in Human Action say nothing that is related to this
issue. In light of the arguments made in Gunning 1998a and 1998b, one must conclude that
Cordato is mistaken.
3. Institutional Structure and Uncertainty
Cordato implies that a judge who follows the Coasean prescription would introduce
uncertainty by failing to enforce already-established entitlements. Referring to Coase’s farmer-rancher case, he suggests that even if property titles are clearly defined so that the farmer has full
rights to the products of his effort on the land, the Coasean judge might consider awarding the
right of use to the rancher, whose cattle strayed onto the land. The Coasean judge, as Cordato
sees it, would balance the benefits against the costs. Cordato objects on the grounds that to shift
an initial entitlement on the basis of benefit-cost analysis would create uncertainty about property
titles that already exist (Cordato 1992: 76-77).
This would hamper the trial and error process in
exchange.
In this part, we shall make three points. The first is that, for the two-person case in which
a right associated with material property had been acquired legally, Coase would not recommend
that judges be given the power to transfer the right from the owner to a non-owner. There are
three reasons. First, transactions costs are low. Second, a judge could hardly calculate the
personal appraisals as well as the individuals themselves could. Third, giving the judge such power
might cause resources to be used merely to attain a favorable court judgment. Thus, for the two
person case, even though the individuals may have to incur transactions costs to bring about the
reassignment of rights that is necessary to “maximize social product,” Coase would surely expect
these costs to be substantially lower than the costs of some other institutional arrangement. And if
the reassignment did not occur because of the high transactions costs, Coase would presume that
this is the best that can be done under the circumstances.
The second and third points require more discussion. The second point is that Cordato’s
argument rightly applies to only a special set of cases. Such cases involve what is typically called
the common property resource. In these cases, there is a genuine possibility that erroneous
Coasean judicial decisions would hamper the trial and error process. However, the situation seems
relatively easy to remedy by confining judicial decisions to “easy cases.” Moreover, if the situation
cannot be remedied, Coase would apparently recommend giving judges the power only if the cost
associated with a hampered trial and error process was less than the benefit due to the decision-making. And he would recognize the difficulty of calculating these benefits and costs. The third
point is that in tort actions, Cordato does not sufficiently support his claim that an alternative rule
would introduce uncertainty. We discuss the second and third points in turn.
The Common Property Resource
Consider a single factory located in a populated valley. Assume that it emits a gas into the
air which, up to now, science has not recognized as having any particular effect. We might even
assume that the gas has gone totally undetected by valley residents, the factory owner, and
science. Now suppose that as a byproduct of some other research, science discovers that the gas
is harmful to residents’ health. In essence, the research discovers a new and unowned factor of
production: the air space that acts as a vehicle for the gas emission. We choose this example to
avoid the complications associated with intentional harm, negligence, and contributory negligence.
We discuss these issues in the next subsection.
We consider two cases. In the first, we assume that the harm is very low relative to the
factory’s expense of preventing the gas from being emitted.
In the second, we assume that the
harm is very high relative to the factory’s prevention expense.
Two characteristics differentiate
these cases from the railroader-farmer case: (1) the effects of the gas are felt jointly by many
valley residents and (2) there are significant costs of transacting, since a collective decision by
valley residents may be required. It is the presence of these transactions costs that, according to
the benefit-cost calculations, may warrant a redistribution of the rights embodied in material
property. Most of our attention will be on the first case since it will, for the most part, be
sufficient to demonstrate the kind of reasoning one must follow to do a proper benefit-cost
analysis in the Coasean sense.
In the first case, there is no problem if the initial property rights are owned by the factory
owner. So let us assume that they are owned by the valley residents. In other words, we assume
that valley residents have the initial right to prevent the factory-owner’s gas emission acts. We
may even assume that the factory owner and the residents previously negotiated an agreement that
resulted in the current ownership. Under the circumstances, it would be efficient in a total product
sense for the right to be transferred to the factory owner, according to our assumptions.
However, because of the high costs to valley residents of reaching an agreement on which price to
charge and how to divide the surplus, it may be impossible for him to purchase the right to emit
the gas.
In this case there is a clear difference between the Coase and the Cordato
recommendation. Whereas Coase would apparently consider allowing a judge to take away the
initial rights of the valley residents on the grounds of raising total product, Cordato would
demand strict compliance with those rights. To have an institutional structure that allows a judge
to take away the initial property rights would introduce uncertainty, he would argue, thereby
compromising catallactic efficiency.
To properly evaluate Coase’s position, we must understand his recommendation in its
totality. Most important, one must realize that Coase recommends taking account of all costs
associated with a proposed change. These include not only the transactions costs, the costs
associated with the harmful effects, and the costs of enforcing compliance with a court order but
also the costs associated with judicial error and the costs due to uncertainty about the future that
the prospect for error leads people to feel. Because Coase includes all costs, his recommendation,
in spirit at least, is not that judges should decide every case. It is, subject to considerations
discussed below, that they should decide easy cases. If a judge believes that his estimate of the
various benefits and costs are likely to be wrong, he should stand back and let the parties work
things out if they can.
It is impossible, of course, to design a system that would completely eliminate the
prospect of judicial error. It is this fact that lends some support to Cordato’s criticism. However,
it does not warrant the strong criticism of Coase that he levels. Nor is the idea that the judge
would have to calculate general equilibrium values especially relevant here. He need only estimate
the value of the right to the factory owner and the sum of the values to the residents. To do this,
he can follow the procedure of allowing them to present their best cases, subject to counter-arguments by the opponent and perhaps their expert economist witnesses. Through all this, one
must keep in mind that the judge has the duty, consistent with the assumption that he will only
take easy cases, to dismiss nuisance claims that he regards as hard cases. If he anticipates that the
costs of making a decision, including the costs incurred by the litigants themselves, are too high;
he should not hear the case. And if he initially decides to hear it but later changes his mind, he
should drop it.
Nor is it correct in light of Coase’s recent reflections on studies of actual situations to
criticize him on the grounds that a system in which judges would only choose easy cases is not
politically feasible. For example, one might claim that the special interest politics inherent in a
democratic system would lead judges who had the discretion to decide which cases are easy and
which are hard to abuse their discretion. If the costs of such a system are expected to be greater
than the benefits -- and Coase might well anticipate that they would be -- then judges should be
prohibited from interfering with the initial rights assignment. In this event, catallactic efficiency in
Cordato’s sense would be achieved.
So would economic efficiency, as Coase uses the term.
We now turn briefly to the second case, in which the external cost to residents is very high
relative to the benefits to the factory owner. We assume that the factory owner has the initial right
to emit the gas, perhaps as a result of an earlier agreement. In a case like this, Coase
recommended in his 1960 “social cost” paper that regulation is a possible solution (Coase 1988a:
116-7). However, emphasis should be placed on the term “possible.” The reason is that, as Coase
later pointed out in a discussion of a possibly different kind of regulation, “[w]hen I wrote my
1960 article...I knew next to nothing about regulation in the United States.” Through his
editorship of the Journal of Law and Economics, he sought to “encourage lawyers and economists
to make...detailed investigations [of government regulation.]” But “[t]he main lesson to be
learned from these studies is clear: They all tend to suggest that the regulation is ineffective or
that, when it has a noticeable impact, on balance the effect is bad, so that consumers obtain a
worse product or a higher-priced product or both as a result of regulation.”(Coase 1996: 107-8)
The message that comes through clearly from Coase is that although a theoretical case can be
made for regulation, the empirical evidence to date suggests that under the conditions that have
prevailed in the past, regulation has not been successful in achieving its intended goal. It follows
that one should be skeptical of recommendations for regulation in the future.
Strict Liability and Uncertainty
Cordato’s argument for strict liability begins by citing the railroad-farmer example.
According to Cordato this example and others like it, although they are concerned with
disequilibrium processes, focus on the equilibrium end state. He writes:
[T]he focus [of the Coasean discussion] is firmly fixed on the general equilibrium end state, which,
implicitly, is part of the ceteris paribus conditions. But in a truly disequilibrium world where the end
state is a continuously moving target, it does not make sense for legal rulings to attempt to insure any
particular market outcome or distribution of resources (Cordato 1992: 98, partly quoted above).
He adds that there is an additional problem of summing up costs and benefits among individuals
(ibid.: 99). He applies this assertion to tort cases in which the title to the property is not clearly
delineated, claiming that “the purely cost-benefit approach of Coase cannot be invoked to solve
the problem”(ibid.).
His only discussion of situations where property rights are not clearly delineated is that of
coming to the nuisance. He writes that although benefit-cost analysis can be no guide, the norm of
“catallactic efficiency can offer some guidance for certain kinds of nuisance problems...”(ibid.:
103). Using the case of an upstream industrial polluter of an unpolluted river, he suggests a first-comer’s rule: “[T]he prima facie case would be in favor of whoever was making use of the stream
first.” In defense of this rule, he writes:
From the perspective of catallactic efficiency, such a rule, consistently applied, would aid both the
market process and the process of individual goal seeking. Its primary efficiency property relates to the
certainty it would provide in the planning and exchange process (ibid.).
In addition, “such a rule would also send important signals to potential ‘comers’ to a nuisance”
(ibid. 104).
He goes on to contrast this approach with that of Donald Wittman, which he considers
representative of Coase. He appears to correctly criticize Wittman for suggesting that institutions
in reality could be set up that would consistently determine the economically efficient sequence of
inputs into a production process. So far as this author is aware Coase himself never suggested this
alternative. What we shall be concerned with here, however, is that Cordato’s case for strict
liability is very weak even on its own grounds.
The discussion in this paper has focused on externality cases in which the medium, or
resource, through which the externality is transmitted has not previously been assigned. In other
words, it has assumed that there is no initial entitlement to the medium. It is evident that the use
of a first-comer’s rule would compromise this assumption. It seems possible in the vast majority
of cases to make a decision on who was a first-comer. Although such a decision would be subject
to political pressure, it is certainly more objective than a decision based on benefits and costs.
However, there are important consequences of applying the first-comer rule that need to be taken
into account. We can understand why this is so by returning to the example of the polluting
factory. We consider two cases. In the first, we assume that the valley residents are the first
comers. In the second we assume that the factory-owner is the first comer.
Valley Residents as First Comers
Suppose that valley residents are the first comers and that there is complete certainty that
the judge would determine this. Under a rule of strict liability, a factory owner would have an
incentive to take into account the full effects of his actions on the valley residents. The cost that
he associates with any action would include not only his production costs but also the anticipated
external costs. Moreover, he would have an incentive to incur costs to discover the external
effects of his actions. Finally, he would have an incentive to anticipate changes in the environment
that would lead any of his actions, according to the “but for” rule, to cause harm to valley
residents.
From the perspective of the efficient use of resources, problems arise when the perpetrator
cannot predict the size of the expected external costs. It is a good thing, from an efficiency
perspective, for the factory owner to take account of the effects of his actions and to attempt to
incorporate the external costs of those effects in his calculations. Even the additional cost that he
incurs of trying to anticipate changes in the external environment that would make him vulnerable
to a damage suit may be good. However, the additional risk he incurs due to changes in the
natural environment that he could not have reasonably been expected to predict is not a good
thing. Because he expects to pay damages in the event of an unanticipated change in the external
environment, he will add a risk cost to his calculation.
From the standpoint of catallactic efficiency, all of these additional costs to the factory
owner reduce the factory owner’s opportunities to participate in the trial-and-error exchange
process. However, they are offset by increased opportunities on the part of valley residents, since
residents need not be concerned about uncompensated harm due to the factory-owner’s actions.
Another point to consider is the absence of any incentive on the part of valley residents to
take avoidance actions. If transactions costs were low, avoidance actions might be spurred by the
factory-owner’s bribe offer. However, the high transactions costs make this infeasible. A judge
could encourage avoidance actions by threatening to fine valley residents who did not take them.
But to do this, the judge would have to study their preferences and alternatives. A rule that
requires this does not seem consistent with Cordato’s arguments about what we can expect
judges to be able to accomplish. Furthermore, if judges did not make any determination at all of
avoidance possibilities, residents might take actions that would increase the size of the damage
claim. They would presumably take any action for which the expected increase in damages was
greater than the personal cost of the action. The relevance of such actions to resource use
efficiency is clearly negative. One can refrain from speculating on their relevance to catallactic
efficiency.
The Factory Owner as the First Comer
Now let us suppose that the factory-owner is the first comer. He would have no incentive
to take account of the effects of his actions on downstream users of water, unless the downstream
users, having learned of the problem, form a collective and offer him a bribe. However, if
transactions costs are high, they could not do this. It is important to apply this analysis to the
situation where the effects of an action are at present unknown yet might be discovered with some
investment. Assume, for example, that the factory owner (and even science itself) is at present
unaware of any downstream effects. The factory owner might be able to discover the harmful
effects at a low cost. However, he would have no incentive to do so. Moreover, since valley
residents were not first comers, they could only use the stream with uncertainty about the future.
They would presumably know that they could never receive compensation for any damage caused
by pollution or any act of nature even if the “but for” test could be applied. In essence, the stream
belongs to the factory owner. It does not matter how long they use it, how reliant they might
become on the stream’s continued absence of pollution, whether the pollution is a side effect or
deliberately intended to extort a bribe, whether the pollution would not have occurred if there had
not been an unanticipatable natural environmental change, and whether the factory owner could
have cheaply avoided causing the effect. The overwhelming consideration is that the factory
owner was a first-comer. The downstream users have a strong incentive to seek out avoidance
techniques.
From the perspective of the efficient use of resources, the factory owner does not
sufficiently account for the external effects of his actions. From the perspective of catallactic
efficiency, the opportunities of the factory owner to participate in trial and error exchange are
higher, while the opportunities of the valley residents are lower.
Broadly speaking, the problem, from the Coasean perspective, seems to be that Cordato’s
strict liability rule, combined with a first-comer’s rule, aims to assign rights to the media through
which external effects are transmitted. However, if these media are common property resources,
high transactions costs make the exchange of the right infeasible in many cases. As a result, in
some cases, enterprise by second comers and avoidance by first comers would be discouraged by
the anticipated high costs of damage judgments; while, in other cases, negligence by first comers
and avoidance by second comers would be encouraged. This means that, from the point of view
of a zero transactions-cost exchange, there would be an inefficient mix of precautionary and
avoidance activities. Whether these apparent inefficiencies of resource use are real inefficiencies
from the positive transactions-cost perspective depends on the existence of a relatively unbiased
court mechanism that can make decisions at a reasonable cost.
The effect on catallactic efficiency is difficult to know since this concept seems incapable
of precise definition.
The Negligence Rule
The alternative to a strict liability rule is a negligence rule. It seems, therefore, that the
target of Cordato’s criticism is a negligence rule, presumably one that would account not only for,
say, the factory owner’s negligence but also for avoidance negligence by the valley residents. It is
important, as Cordato recognized, not to confuse the use of a negligence rule in externality cases
with its use in various consumer products cases.(Cordato 1992: 101) In the latter cases,
proponents of a negligence rule aim to deny individuals the right to make enforceable contracts. It
is difficult to conceive of realistic circumstances under which someone who emphasized efficiency
in resource use would endorse a rule denying such a right.
Negligence as it applies to the case of the perpetrator of an externality refers to the
perpetrator’s right to avoid being held liable if he can prove that he was not negligent. In effect,
this means that he can claim that the cause of the externality was beyond his control. He can claim
that there was no reasonable way for him to find out in advance (1) either that his action would
cause the harm or (2) the high amount of harm that it caused. From the viewpoint of the recipient
of the external effect, the negligence rule means that she must prove either intent or negligence on
the part of the perpetrator in order to claim damages.
Cordato’s discussion of a judge’s calculation problem seems to be concerned with the
problem of determining two things: (1) the loss caused by the action and (2) the cost of the
precautions that the perpetrator could have taken. However, he does not distinguish the two. He
appears to simply claim that no judge could calculate these values in a dynamic setting.
However, it must be noted that at least one half of the problem still exists in the case of a strict
liability determination. The amount of damages must still be determined. The only thing that the
negligence rule adds is the problem of determining the costs faced by the perpetrator of finding
out about the harmfulness of his effects and of taking proper precautions. Interestingly, Cordato
apparently admits as much (ibid.: 105-6). Yet he does not see the relevance of this to his earlier
argument.
It is the difficulty of determining intent and negligence that might lead one to estimate that
the uncertainty costs of using a negligence rule would outweigh the benefits. Under strict liability,
intent and negligence are irrelevant, so no court case resources are spent on determining these
states of mind. They are devoted entirely to determining the extent of damages. Under the
negligence rule, intent and negligence are in the forefront. If it is determined that there is no intent
and no negligence, damages need not be calculated at all. If it is determined that there is intent or
negligence, these states of mind need to be determined in addition to the damages. Whether
Cordato’s realizes this difference is uncertain.
4. Conclusion
Although Cordato’s Austrian welfare economics has not fared well under closer scrutiny,
the exercise has served to clarify several important issues. The aim of this conclusion is to make a
list of these. First, any effort to identify an ideal institutional structure for the operation of the
market economy ought to be based not on the concept of real property but on the concept of
factors of production. What is important to an individual in his exercise of freedom of choice is
anything that he believes exists in a causal connection with the satisfaction of his wants. In other
words it is anything that he regards as a factor of production. Similarly, what is important to the
consumers in the market economy is that individuals acting in the role of entrepreneurs have the
freedom to own, invent and use anything that they regard as factors of production. To base an
ethical system, or a welfare criterion, on real property is to commit the materialistic fallacy.
Second, Coase should not be classified as a naive interventionist. Not only has he
demonstrated his appreciation for the essentials of the Austrian critique of central planning, he has
not recommended intervention in practice. Like any good economist, he has recognized that, from
the standpoint of consumer want satisfaction, a theoretical case can be made for market
intervention. Like Mises the economist, he has explored this (in his assessment of Pigou and the
Pigovian tradition) by asking whether interventionist proposals will actually achieve the results
that the proposers want. His answer differs from that of his predecessors in that he has recognized
ways to exchange that were ignored or disregarded. Real transactors in the market economy can
often find solutions to externality problems that are not evident to Pigovian policy-oriented
economists. Because the Pigovians did not realize this, they were inclined to overlook the
alternative of doing nothing and they were prone to err in predicting the effects of their policy
recommendations. Moreover, they neglected the costs associated with carrying out the
intervention itself. The fact that Coase used the competitive equilibrium model to help him make
these points marks him not as an interventionist but as a competent expositor of his ideas.
Third, this paper has shown that Cordato’s case for strict liability based on uncertainty is
at best weak. Since he appears to have neglected the common property resource, which is
typically the medium through which external effects are transmitted, his case even appears
somewhat naive. Yet, his stand against the negligence rule and Coasean analysis was extremely
critical and dismissive. To this author, there appear to be two requirements that would have to be
met in order to strengthen his apparently weak case. First, he would have to define catallactic
efficiency more precisely. Most importantly, he must discuss situations in which an increase in the
freedom to pursue goals through a trial and error process by one person might be accompanied by
a decrease in the freedom of others. Second he would have to make a tight comparison of cases.
It is conceivable that in such a comparison, the strict liability rule would win out. But there is no
way to know from the analysis he has given us so far.
References
Coase, R. H. (1937) "The Nature of the Firm." Economica 4: 386-405. Reprinted in R. H. Coase
(1988a) The Firm, the Market, and the Law. Chicago, Ill.: University of Chicago Press.
Coase, R.H. (1960) "The Problem of Social Cost." The Journal of Law and Economics. 3: 1-44.
Reprinted in R. H. Coase (1988) The Firm, the Market, and the Law. Chicago, Ill.: University of
Chicago Press.
Coase, R.H. (1972) Comment on “The Muted Voice of the Consumer in Regulatory Agencies.”
In W. J. Samuels and H. M Trebing (eds). A Critique of Administrative Regulation of Public
Utilities. East Lansing, MI: Institute of Public Utilities, Graduate School of Business
Administration, Michigan State University.
Coase, R. H. (1974) “Economists and Public Policy.” In J. F. Weston (ed.). Large Corporations
in a Changing Society. New York: New York University Press.
Coase, R. H. (1988a) The Firm, the Market, and the Law. Chicago, Ill.: University of Chicago
Press.
Coase, R. H. (1988b) Blackmail, Virginia Law Review. 7 (4): 655-76)
Coase, R.H. (1996) “Law and Economics and A.W. Brian Simpson.” Journal of Legal Studies.
25 (103-119)
Cordato, Roy E. (1992) Welfare Economics and Externalities in an Open Ended Universe: A
Modern Austrian Perspective. Boston: Kluwer Academic Publishers Group.
Cowen, Tyler. (1991) "What a Non-Paretian Welfare Economic Would Have to Look Like." In
Don Lavoie (ed.). Economics and Hermeneutics. London: Routledge.
Endres, A. M. (1997) Neoclassical Microeconomic Theory. New York: Routledge.
Gunning, J. Patrick (1998a). "Austrian Welfare Economics?: A Misesian Response." (manuscript)
Gunning, J. Patrick (1998b). "The Liberal Economist's Dilemma: Rothbard's Critique of Mises's
Value-Free Economics." (manuscript)
Hayek, F. A. (1976). Law, Legislation, and Liberty, Vol. 2: The Mirage of Social Justice.
Chicago: University of Chicago Press.
Herbener, Jeffrey M. (1997) “The Pareto Rule and Welfare Economics.” Review of Austrian
Economics. 10 (1): 79-106.
Kirzner, Israel M. (1992) "Welfare Economics: A Modern Austrian Perspective" in Israel M.
Kirzner (ed.) The Meaning of the Market Process: Essays in the Development of Modern
Austrian Economics. London and New York: Routledge. 180-192.
Rothbard, Murray N. (1956) "Toward a Reconstruction of Utility and Welfare Economics" In
Mary Sennholz (ed.) On Freedom and Free Enterprise. New Jersey: D. Van Nostrand. Reprinted
in Rothbard (1997) The Logic of Action I: Method Money, and the Austrian School: Cheltenham
UK: Edward Elgar.
von Mises, Ludwig (1966). Human Action: A Treatise on Economics. Chicago: Henry Regnery
Company.
von Mises, Ludwig (1978) Liberalism. Kansas City: Sheed, Andrews and McMeel.
Gunning’s Address
J. Patrick Gunning
Visiting Professor
U.S. Coast Guard Academy
Management Department
15 Mohegan Avenue
New London, CT 06320
Please send feedback:
Email: gunning@nomadpress.com
Go to Pat Gunning's Pages