M PRA

Munich Personal RePEc Archive

The Social Costs of Monopoly: A Survey

And An Evaluation
Erdal Gumus
Eskisehir Osmangazi University

2006

Online at https://mpra.ub.uni-muenchen.de/42107/
MPRA Paper No. 42107, posted 21 October 2012 18:03 UTC

Akdeniz ..B.F. Dergisi (13) 2007, 149-164

THE SOCIAL COSTS OF MONOPOLY: A SURVEY AND AN

EVALUATION
MONOPOLN SOSYAL MALYETLER: LTERATR TARAMASI
VE BR DEERLENDRME
Erdal GM
ABSTRACT
Existence of monopoly and its costs to societies have been intensely
studied. However, there has been no clear view obtained yet. Studies of rentseeking approach of the public choice school even make the issue more
interesting. This paper surveys from traditional Harbergers triangle
approach of social costs of monopoly to the most recent studies and drive
some conclusions from it.
Keywords: Social costs of monopoly, Harbergers triangle, Rent-seeking theory.
ZET
Bir ekonomideki monopolc firmalarn varl dolays ile sz konusu
firmalarn topluma ykledikleri maliyetler youn bir ekilde aratrma konusu
yaplmaktadr. Ancak, konu zerinde gr birlii henz ortaya kmamtr.
Kamu tercihi okulu rant arama yaklamnn almalar konuyu ok daha
ilgin hale getirmektedir. Bu alma ile monopoln sosyal maliyetlerini
Harbergerin gen yaklamndan, yakn zamana kadar gelitirilen
almalarn bir literatr taramas yaplarak deerlendirilmekte ve baz
sonulara varlmaktadr.
Anahtar Szckler: Monopoln sosyal maliyetleri, Harberger geni, rant arama
teorisi.

Eskiehir Osmangazi niversitesi BF Maliye Blm retim yesi

The Social Costs Of Monopoly: A Survey And An Evaluation

INTRODUCTION
The social costs of monopoly have been an interesting subject in the
literature since Harbergers study in 1954. Since then, many scholars such as
Stigler (1956), Tullock (1967), Posner (1975), Wenders (1987), Browning 1
(1997), Epstein and Nitzan (2002), Brown and Yoon (2006), and others 2
have studied various aspects of the issue. Many of them have tried to
measure empirically how large the social costs of monopoly might have
been while others have studied the theoretical aspects of this subject.
In an economy where a monopoly is present, there is a cost to society,
since a monopoly firm sets its output price above the competitive
equilibrium price level and its output level would be lower than that of
competitive level. Hence, there will be excess demand that would not be
fulfilled, and therefore society suffers from that gap too.
The purpose of present study is to survey the literature briefly and
evaluate them and drive some results. The paper is organized as follows:
Current partial equilibrium research results are summarized in the next
section. From Harbergers study to most recent studies that employ general
equilibrium methodology have been surveyed in section III. Section IV
hosts an evaluation and conclusion.
WELFARE LOSS
Although many early studies on the social cost of monopoly power
have been developed and analyzed the issue in the context of the
conventional partial equilibrium framework, there have been recent studies
that started to evaluate this issue from general equilibrium perspective.
The social cost of monopoly was methodologically introduced by
Harberger (1954) as the welfare triangle, ABC, which is called in the
literature as deadweight loss, and Harberger triangle, or welfare loss,
that is shown in figure I.

Browning (1997) offers a new factor in calculating social cost of monopoly. His
suggestion is to include resource supply distortion effect caused by tax system.
Taking tax distortion into consideration, his findings indicate a huge social cost of
monopoly when comparing with the conventional estimate.
2 Worcester (1973), Bergston (1973), Cowling and Mueller (1978), Baik (1999),
Davis and Reilly (2000) can be mentioned here.
1

150

Erdal GM

Pm

Pr

Pc

III

II
IV

MC

B
D
Qm

Qr

Qc

Quantity

Source: Modified from Wenders (1987, 458).

Figure 1: Harberger Triangle

Harberger states that when monopoly price is above competitive level

there would be a loss to society since buyers would buy its product (Qc-Qm)
if it would produce at the competitive level. Thus, in the static framework,
the area of ABC in figure I may constitute the social cost of monopoly. The
area PmABPc is excess profit which is a transfer from consumers to
monopolist (Harberger, 1954).
Whether the social cost of monopoly is just the said triangle shown
in figure I has been the subject of many studies. Many of these studies have
relied on the simplicity of partial equilibrium framework while many others
have adopted a powerful computable general equilibrium theory to analyze
and hopefully find relatively accurate results. In the rest of this paper, we try
to be selective and review those important studies adding a positive aspect
to the issue.
Tullock (1967) and Posner (1975) have extended the analysis of welfare
loss from monopoly by considering the dynamics of monopolization as
Baysinger and Tollison mentioned (1980). Tullock and Posner argue that
The existence of an opportunity to obtain monopoly profits will attract
resources into efforts to obtain monopolies, and the opportunity costs of
those resources are social costs of monopoly too (Posner, 1975:807). This
statement simply says that potential monopolists will expend resources on
monopolizing inputs, which, as Baysinger and Tollison (1980) mention,
would be in forms of hiring lawyers, lobbyists, improved product quality,
151

The Social Costs Of Monopoly: A Survey And An Evaluation

better services etc., up to the value of area I and III or rectangle of PmPcBA
in figure I in order to acquire the rights of monopoly rents. They especially
point out that since the process of monopolization is a transfer between
members of society, any resources expended in this effort are socially
wasted. Moreover, the social loss from monopoly, according to them, is not
simply the area of the ABC triangle, but the entire trapezoid, PmACPc, as
shown in figure I.
Wenders has argued that Recurring or sunk, even the largest
specification of the Harberger and Tullock costs of regulatory
monopolization may fall far short of the actual welfare costs. This is because
the analysis concentrates on the rent-seeking Tullock costs and largely
ignores the parallel rent-defending 3 Tullock costs. A proper assessment of
such rent-defending Tullock costs might more than double the maximum
welfare costs of regulation suggested by Posner (Wenders, 1987:456). This
means that consumers will also actively defend consumer surplus, and they
will not watch when monopolists are rent-seeking. Rather they will parallely
take actions against monopolists. This also requires some consumer
spendings such as lobbying action, or some types of pressure on regulatory
agencies to force monopolists to set prices as closely to competitive levels as
possible. Wenders (1987) explains the reasons that
Neither buyers nor sellers may refrain from spending the maximum
amount they each have at stake. If either voluntarily spends less, they will be
taken advantage of by the other side. In addition, there is the incentive for
each side to compete among themselves to either achieve or avoid the
proposed regulation. Thus, both sides may spend up to the amount each has
at stake (Wenders, 1987:458).
In figure I, consumers will spend as much as the area of PmACPc to
pay the price of Pc for the good in question. In this sense, Wenders claims
that in a situation where monopolization is not binary, sellers are suggesting
a form of regulation that would result in a Pm price, while buyers have the
alternative of engaging in rent-defending activities that would hold the
regulated price below Pm, say, at Pr, which might lie anywhere between Pm
and Pc. This follows that the sellers initially propose a regulation that would
result in price Pm, the full monopoly price. Buyers would be willing to
spend an amount up to I+II to water down the proposed regulation so only
price Pr came about. For the water down regulation, sellers would be willing
to pay III + IV in figure I. Hence the total welfare costs might amount to
the entire trapezoid between Pm and Pc that is I + II + III + IV + V
(Wenders, 1987:458). Consequently, Wenders concludes that when rent3 Perhaps a more descriptive term would be consumer surplus defending [
Wenders footnote.]

152

Erdal GM

defending Tullock costs are admitted to the analysis, and presuming a

parallel kind of perfect dissipation of the consumers surplus lost to full
monopoly pricing, the total welfare costs amount to a similar, but
necessarily larger, Posner-like trapezoid computed at the unobserved full
monopoly price (Wenders, 1987:458).
This conclusion, of course, is not the final conclusion. There are other
theoretical aspects of the issue such as price elasticity of demand for goods
in question which is also important when studying empirical aspects of the
subject. However, we will review those aspects of the issue in the next
section. Here, it can be said, in sum, that the social costs of monopoly
greatly vary depending upon the size of relevant elasticities, the distribution
of monopoly markups, and of course, the size of the monopolist sector in
the economy. Therefore, we agree with Wenders who expresses rightly that
At one time, the cost of monopoly was neatly settled and made its way into
even elementary texts. By now, it should clear that these matters are far
from settled. A new, stable, orthodoxy in the area is still a long way off
(1987).
THE LITERATURE
Harbergers 1954 seminal study is one of the first influential works that
dealt with social cost of monopoly problem. In his study, Harberger (1954)
emphasizes about the measurement of resource misallocation and welfare
loss due to monopoly. Based on a number of important assumptions 4, he
uses Epsteins (1934) study as a source of data and calculates rates of return
for each industry for the five year period, 1924-1928. 5
Harberger estimates welfare loss of monopoly in the amount of less
than one-tenth of one percent of the national income (Harberger, 1954:82).
Figure II reproduces Harbergers original figure that shows Welfare Loss
or called Harberger triangle in the literature. His main conclusion was that
all I want to say here is that monopoly does not seem to affect aggregate
welfare very seriously through its effect on resource allocation (Harberger,
1954:87).

His assumptions are (1) in the long run, resources can be allocated among our
manufacturing industries in such a way as to yield roughly constant returns.
(Harberger, 1954:77), (2) All firms are operating on their long run cost curves, the
cost curves are so defined as to yield each firm an equal return on the invested
capital, and markets are cleared (Harberger, 1954:78), and, (3) The elasticity of
demand for the industrys product is unity (Harberger, 1954:78).
5 Epstein, Ralph C., Industrial Profits in the United States (National Bureau of
Economic Research, 1934).
4

153

The Social Costs Of Monopoly: A Survey And An Evaluation

This conclusion is very important in the sense that after this study,
many scholars have questioned both sides of his study; theoretically and
empirically. Another feature of his study is that it was one of the first
comprehensive studies dealing with monopoly and resource allocation.
The weak point in Harbergers study is his assumptions on which
Stigler (1956) raises several objections. Stigler criticizes Harbergers
assumptions and says the assumption of a demand elasticity of unity is
objectionable. A monopolist does not operate where his marginal revenue is
zero (Stigler, 1956:34). Thus, he says in Harberger model, welfare losses go
up when elasticity of demand increases (Stigler, 1956:34).
Price,
Cost

"Welfare Loss"

Excess Profits
Unit Cost
Incremental
Resources

Demand

Quantity

Source: Reproduced from Harberger, 1954, 78)

Figure 2: Traditional Harbergers Triagle

The second objection Stigler raises is that If monopoly profits are

capitalized, the capital of a monopolist grows in such a way that only
competitive rates of return are being earned (Stigler, 1956:34). Here Stigler
indicates his objection by saying that some industries in Harberger study are
monopolistic and some are not. In his words The estimate of welfare
losses is biased downward on this score. One can not believe that
petroleum, meat-packing, railway equipment, and pianos were excessively
competitive industries and that boots and shoes, planning mills, canned
goods, and job printing were monopolistic (Stigler, 1956:35). He finds
Harberger calculation of welfare loss due to monopoly too low (Stigler,
1956:34).
Stiglers study encouraged new studies on both sides and two views
were developed on the social costs of monopoly. One of them reflects
Harbergers view that monopoly is not an important problem since the costs
154

Erdal GM

it causes for society are too low, while the other view supports Stiglers
position that those costs might be large.
For instance, Schwartzman (1960) tries to present some figures to
confirm Harbergers finding. He follows Harbergers procedure but uses his
own estimation of profits to measure the welfare cost of monopoly. He
relaxes Harbergers unit elasticity assumption by assuming a range of
reasonable values for the elasticity of demand (Schwartzman, 1960:627). The
welfare loss he calculated for the United States manufacturing industry for
the year of 1954 was $ 202.5 million. 6 The Harbergers calculation for the
year of 1953 was $225 million. Both figures were less than one percent of
the corresponding years national income (Schwartzman, 1960:630). His
conclusion is that Since the estimates are similar to Harbergers, they
provide confirmation of Harbergers general conclusion that the welfare loss
from monopoly has been small (Schwartzman, 1960:630).
Kamerschen (1966) argues that the social costs of monopoly were
much larger than what Schwartzman and Harberger found. Kamerschen
agrees with Stigler on the fact that the welfare loss would be larger than
what has been estimated. In this context, Kamerschens study is an
application of Harbergers technique, and is more comprehensive. In the
light of previous studies in this area, he makes some further realistic
assumptions and uses various values of price elasticity of demand and
concludes that
Estimating by a number of alternative methods, we calculate welfare
losses that range from 1 to 8 percent of the average national income.
Therefore, this study suggests that the welfare losses in the American
economy are of a significantly different magnitude from what previous
studies had indicated and are of a higher order of magnitude. Using what
appear to be most realistic estimates-based on after-tax income, fully
adjusted profit rates with industry-by-industry elasticity data- we obtain
losses of roughly 6 percent of the average national income (Kamerschen,
1966:235).
This conclusion is important when we look at the previous empirical
studies. There is an important increase in the amount of welfare loss
calculation. This is because every study adds positive contribution to the
subject along with available data.
Tullocks (1967) study is a good example of an important theoretical
contribution to the field after Harbergers study. Tullock advocates that the
The original calculated welfare loss was $ 234 million, however, the author of the
article has made a correction in his later article: The effect of Monopoly; A
Correction, Journal of Political Economy, October 1961, pp. 494.

155

The Social Costs Of Monopoly: A Survey And An Evaluation

social costs that monopoly causes are much larger than the Harberger
triangle. His contribution to the subject briefly is that
The rectangle to the left of the welfare triangle is the income transfer
that a successful monopolist can extort from the customers. Surely we
should expect that with a prize of this size dangling before our eyes,
potential monopolists would be willing to invest large resources in the
activity of monopolizing. In fact the investment that could be profitably
made in forming a monopoly would be larger than this rectangle, since it
represents merely the income transfer. The capital value, properly
discounted for risk, would be worth much more. Entrepreneurs should be
willing to invest resources in attempts to form a monopoly until the
marginal cost equals the properly discounted return. The potential
customers would also be invested in preventing the transfer and should be
willing to make large investments to that end. Once the monopoly is
formed, continual efforts to either break the monopoly or muscle into it
would be predictable. Here again considerable resources might be invested.
The holders of the monopoly, on the other hand, would be willing to put
quite sizable sums into the defense of their power to receive these transfers
(Tullock, 1967:231).
He gives an example that the cost of a football pool is not measured
by the cost of the winners ticket, but by the cost of all tickets. Similarly, the
total costs of monopoly should be measured in terms of the efforts to get a
monopoly by the unsuccessful as well as the successful (Tullock, 1967:232).
As his important contribution to the subject he concludes that the resources
put into monopolization and defense against monopolization would be a
function of the size of the prospective transfer. Since this would be
normally large, we can expect that this particular socially wasteful type of
investment would also be large. The welfare triangle method of
measurement ignores this important cost, and hence greatly understates the
welfare loss of monopoly (Tullock, 1967:232).
Tullocks work may be thought as an important contribution to the
subject in two respects; (1) it considers the rectangle to the left of the
Harbergers triangle, PmABPc in figure I, as social cost of monopoly in
addition to triangle. (2) It opens an argumentative door that questions
whether the efforts of potential monopolists and consumers can be
regarded socially wasteful and, therefore, to be added to the social cost of
monopoly.
It is especially interesting to mention Worcester (1973) study that
follows a similar analogy of Harbergers study (1954) and claims that the
present study attempts to reveal monopoly loss which in other studies was
hidden in industries which include both profitable and unprofitable firms
by using data presented by Fortune Magazine for the 500 largest industrial
156

Erdal GM

firms in the U.S. for each year, 1956 through 1969 (Worcester, 1973:234).
Thus he claims courageously that previous estimations of social costs of
monopoly may have been underestimated. However, he concludes The
estimates of welfare loss to monopoly reported here are substantially lower
than anticipated when this research was undertaken. ...I can only conclude
that there is very little ground for the common belief that a large loss of
welfare exists due to the economic impact of monopoly power. as a social
scientist I believe that the burden of proof now rests on those who believe
the loss is large (Worcester, 1973:244).
An empirical study similar to the traditional Harbergers method was
done by Siegfried and Tiemann (1974). They apply a strictly partial
equilibrium framework and conclude that The results of our analysis show
that the bulk of the welfare loss due to monopoly is concentrated in
relatively few industries (Siegfried and Tiemann, 1974:190-191). They
suggest that a policy to reduce monopoly power in specific industries may
possess a greater benefit/cost ratio than the benefit/cost ratio for
restructuring the whole manufacturing sector. Marginal adjustments in
market structure may be desirable even if a total restructuring is
unwarranted (Siegfried and Tiemann, 1974:191).
Posners (1975) study can be viewed as one of the most important
contributions to this literature. Posner (1975) agrees with Tullock (1967) on
the measurement of social cost of monopoly which is not simply
Harbergers triangle but the rectangle to the left of the welfare loss triangle
too. In his words the existence of an opportunity to obtain monopoly
profits will attract resources into efforts to obtain monopolies, and the
opportunity costs of those resources are social costs of monopoly too
(Posner, 1975:807). Posner makes three basic assumptions and constructs a
model to calculate the social cost of monopoly. He states firms expenditures
will be as much as total monopoly profit that
If 10 firms are vying for a monopoly having a present value of $1
million, and each of them has an equal chance of obtaining it and is risk
neutral, each will spend $100,000 (assuming constant costs) on trying to
obtain monopoly. Only one will succeed, and his costs will be much smaller
than the monopoly profits, but the total costs of obtaining monopolycounting losers expenditures as well as winners- will be the same as under
certainty (Posner, 1975:812).
He separates monopolies as regulated monopolies and private
monopolies, and especially stresses that social costs of a regulated monopoly
probably exceed the costs of private monopoly (Posner, 1975:818-819). His
main conclusion is that previous studies of the costs of monopoly may have
grossly underestimated those costs, and the costs of monopoly are quite
157

The Social Costs Of Monopoly: A Survey And An Evaluation

probably much greater in the regulated than in the unregulated sector of the
economy (Posner, 1975:821).
Cowling and Mueller (1978) raise several objections against the
Harberger approach. They use partial equilibrium analysis and estimate
welfare loss from monopoly using procedures derived to meet their
objections (Cowling and Mueller, 1978:727). They state that all previous
estimates of monopoly welfare losses suffer in varying degrees from the
same biases incorporated in Harbergers original estimates (Cowling and
Mueller, 1978:727). After they explain four substantive criticisms of the
Harberger approach and follow Tullocks (1967) and Posners (1975)
methods, they calculate the welfare loss by measuring at the firm-level for
both the United States and United Kingdom. They conclude that our
figures and supporting analysis further demonstrate that the monopoly
problem is broader than traditionally suggested. A large part of this
problem lies not in the height of monopoly prices and profits per se, but in
the resources wasted in their creation and protection (Cowling and Mueller,
1978:744).
This conclusion, driving from empirical estimation, supports what
Tullock (1967) and Posner (1975) advocated. Additionally they suggest,
contrary to those who argue that the social cost of monopoly is small, thus
the monopoly problem is unimportant, that still further weight would be
added against the position that monopoly power is unimportant if the link
with the distribution of political power were considered (Cowling and
Mueller, 1978:746).
Their work was an application based on Tullocks (1967) and Posners
(1975) definitions, and their finding confirms that social costs of monopoly
in fact are considerably higher.
While various explanations and reasoning have been offered and
defended on the theoretical ground of determining the social cost of
monopoly 7, others try empirically to test and come up with hopefully
supportive results. Dixon, Gunther and Mahmood (2001) have estimated
the welfare loss of monopoly in Australian manufacturing. They follow
Cowling and Muller (1978) method and estimate the deadweight loss. They
modified Cowling and Muller model by including the presence of collusive
oligopoly effect in their model. Based on the estimation, they arrived at the
conclusion that the social costs of monopoly in Australia are substantially
higher than previous estimates have suggested (Dixon, Gunther and
Mahmood, 2001:396). Their estimated aggregate welfare loss was 1.38 per
cent of average GDP over the period measured in current price.
See for instance Young (1997) who approaches this issue by studying price
leadership behavior.

158

Erdal GM

Another study confirming Tullock and Posner studies were developed

by Baysinger and Tollison (1980). They extend Tullock (1967) and Posner
(1975) methods by incorporating certain other aspects of reality into the
analysis of the social costs of monopoly and regulation. When the
institutional environment within which potential monopolists and consumer
forces attempt to capture portions of consumer surplus is taken into
account, results at variance with those of both Harberger and Tullock and
Posner are generated (Baysinger and Tollison, 1980:22). They study the past
and current price behavior of regulators under conditions of when state
action is arbitrary and when state action is predictable. Their conclusion is
that the social cost of rent seeking is a variable which is functionally related
to both the past and current behavior of price-setting regulators. Past
behavior is important in that it informs the formation of expectations by
those intimately affected by the regulatory process. The expectations so
formed determine the individually optimal level of resources the parties will
devote to the monopolization process. Current regulatory behavior is
important in that it determines the static allocative losses which must be
added to monopolization costs of monopoly and regulation. It stands to
reason that attempts to extract rents will be fought by affected parties,
unless such a contest is deemed futile due to the past history of the involved
authorities. We have demonstrated that Posners estimate of social waste is
too high if parties can predict the outcome of monopoly rights proceedings
with near certainty, and too low if these outcomes are relatively
unpredictable. thus while the conventional result that rent-seeking
expenditures are socially wasteful stands, the degree of such welfare losses is
related to the nature of the institutional environment within which rentseeking takes place (Baysinger and Tollison, 1980:26).
Therefore, contribution of this study, it can be said, to take into
consideration the past and current behavior of price setting regulators when
estimating the social costs of monopoly.
Baik (1999) sets up a model in which the government decides first
whether to regulate a monopoly or not. If regulation is appropriate, then,
which firm should be the monopolist? In this model Baik (1999) seeks to
address whether consumer surplus defending activities increases or
decreases the social costs of monopoly. His main conclusion is that given
just one rent-seeking firm, consumers CS-defending activities generally
increase the social cost of monopoly, but given two or more rent-seeking
firms, such activities generally reduce the social cost of monopoly (Baik,
1999:551).
Stegemann (1984) also develops a model which investigates foreign
transactions dimension to the issue. His study can be viewed as an
extension of the Tullock, Posner, Cowling and Mueller perspective that
159

The Social Costs Of Monopoly: A Survey And An Evaluation

analyzes the issue in an open economy. In other words, it links the existing
literature to international markets. Stegemann explains the aim of his paper
as ...most national markets are linked, via imports or exports, to
international markets. Of course, everybody is aware that the market power
of domestic producers often depends on the existence of import barriers
and that the magnitude of the monopoly burden may depend on the level of
such barriers. Yet it seems necessary to explore more fully the implications
for the social costs of domestic market power resulting from the existence
of international competition (Stegemann, 1984:718-719). He states that in
open economies, domestic producers can retain monopoly power for three
reasons (1) the domestic market is protected; (2) the supply of importable
perfect substitutes is not perfectly elastic; (3) imports are imperfect
substitutes of domestic goods (Stegemann, 1984:718). He employs partial
equilibrium analysis to investigate the implications for the welfare burden of
monopoly resulting from the existence of foreign competition in each of the
three cases (Stegemann, 1984:718). He concludes that as a consequence,
monopoly in an open economy may cause a welfare burden by causing
excessive importation. Importation is excessive because monopolistic
pricing behavior of domestic producers induces buyers to purchase imports
when equivalent domestic goods could be produced at a lower social
opportunity cost (Stegemann, 1984:719). Therefore, the contribution of
this study is appreciable since it analyzes the foreign sectors effect in the
context of the existence of literature in this area.
Davis and Reilly (2000) use analytical and experimental methods to
evaluate rent-seeking and rent-defending behavior of having a monopoly.
By using various experiments they conclude that overbidding is persistent
when bidders have different sharing rules. In fact, the observed social costs
of rent-seeking often increase just when rent-defending has the greatest
predicted ameliorative effect (Davis and Reilly, 2000:389).
Another good contribution to this literature was developed by Wenders
(1987). His study mainly extends Tullocks study considering the parallel
Tullocks rent-seeking procedure which briefly says consumers would
perform some activity to defend consumer surplus. When considered
simultaneously, potential monopolists would do some activity (e.g.,
lobbying, hiring lawyers, giving bribery, ect.) to get consumer surplus as
monopoly rent, while consumers would do parallel activity not to surrender
consumer surplus to monopolists. It is from in this context that Wenders
concludes
Under perfect competition among and between buyers and sellers,
regulation always results in maximum welfare costs, that is, equal to the
rent-seeking Tullock and Harberger costs under full monopoly pricing. And
only if sellers succeeded in attaining full monopoly pricing would this result
160

Erdal GM

equal the total welfare cost originally proposed by Posner; in all other cases
the total welfare loss would be larger due to the admission of rent-defending
Tullock costs (Wenders, 1987:458).
Studies considering consumer defensive activities against monopoly
power give conflict results. Epstein and Nitzan (2002) find that social costs
of monopoly would not be large while Keem (2001) strongly reveal that, by
assuming asymmetric lobbying abilities across firms, consumer defensive
activities will increase the social cost of monopoly power.
A recent empirical study (Yoon, 2004) takes the traditional partial
equilibrium approach and uses company data in calculating welfare loss of
monopoly in Korean economy. This study uses manufacturing data of
selected Korea companies and found 7.56 percent of gross value of
shipments in 1998 as costs to society resulting from monopoly power in
Korea (Yoon, 2004: 956).
Given that partial equilibrium analysis isolates consumers utility levels
and states firmly that monopoly firms seek only to maximize their profits
leaded researchers to consider the social cost of monopoly in general
equilibrium framework. Bergson (1973) analyzed the subject in the general
equilibrium framework to prove that the social costs caused by monopoly
are large. Bergson, after Stigler, seriously criticized the partial equilibrium
framework that mainly undergirded Harbergers approach of a general
equilibrium model. As Worchester (1975) states, Bergson uses a constant
elasticity of substitution (CES) utility function and a linear production
function to calculate the compensating variation -the additional income
consumers must receive to offset the loss of utility caused by monopoly
price. The share of the total market monopolized, the size of monopoly
prize markup over competitive price and different elasticities of substitution
(used in place of demand elasticity), are built into his model. Potential
welfare losses are found by assuming specific values for the parameters. He
then offers two tables showing welfare losses calculated for a number of
hypothetical alternatives. He also advocates strongly that monopoly can be a
matter of consequences (Bergson, 1973). In sum, his contribution on
theoretical ground is useful for further research; however, the parameters in
his tables are questionable.
Recently the general equilibrium model applicable to measure social
cost of monopoly was reconsidered in context of cost minimizing market
equilibrium developed by Brown and Wood (2004a; 2004b) and later
improved by Brown and Yoon (2005; 2006). Brown and Wood (2004a) have
argued that the purpose of regulating monopolies is to promote allocative
efficiency. The deviation of allocative efficiency measured by Debreus
(1951) coefficient of resource utilization can be defined as waste resources
that represent as social cost of monopoly power. Thus, they developed a
161

The Social Costs Of Monopoly: A Survey And An Evaluation

general equilibrium model where monopoly firms are assumed both to be

price takers in competitive factor markets and quantity takers in their output
markets (Brown and Wood, 2004a). They have illustrated their cost
minimizing general equilibrium model by a number of numerical examples
and later revised their work (Brown and Wood, 2004b) by following study
of Shoven and Whalley (1992) to reformulate their general equilibrium
model. Despite their rationality, they calculated upper and lower bounds on
the margin between the monopoly price and the contrafactual competitive
price in each observation (Brown and Wood, 2004b, 13). Brown and Yoon
(2005; 2006) have followed and confirmed Brown and Wood (2004b) study
on theoretical ground.
EVALUATION AND CONCLUSION
In this paper, we have surveyed many important studies that employ
both partial and computable general equilibrium models to develop
theoretical bases of a comprehensive measurable social cost of monopoly.
Most of the studies relied on conventional partial equilibrium model while
recent studies prefer applicable general equilibrium analysis.
Based on these models, there have mainly been two different views
developed in the literature to estimate the welfare effect of the monopoly.
One view is being that the social costs of monopoly are just Harbergers
triangle, and therefore, they are negligible since they are considered to be
relatively low. The second view is being that the dynamic of monopolization
should be considered, and the opportunity costs of all efforts by all relevant
sides in trying to capture and protect monopoly rents should also be
considered wasteful, and therefore, those costs of efforts should well be
considered as social costs of monopoly too. When the latter view is
accepted in its larger meaning, the social costs of monopoly might be
enormous. On the theoretical ground the literature seems to value the latter
view whilst empirical studies do not give relatively strong support to this
view. There are many factors that affect the relative size of the social cost of
monopoly. They can be listed as follows;

162

Assumptions used

Effect of tax distortion

Behavior of regulating authority

Relative size of rent protecting activities

Relative size of rent defending activities

Degree of economic development of the economy

Degree of differentiation of price from marginal cost

Erdal GM

Relative size of monopolized industry in the economy

Price elasticity of demand in the monopoly industries

Relative size of consumer surplus defending activities

On empirical grounds, the estimated social costs of monopoly have
been regarded as low due to technical and measurement difficulties to grab
all relevant factors in to the analyses. Thus, no one can still claim that the
size of social costs of monopoly truly estimated, however, new computable
general equilibrium approaches seem promissory and future research may
concentrate on this theory to improve it and come up with much better
understanding.
REFERENCES
BAYSINGER, B. AND TOLLISON R. (1980) Evaluating the Social Costs
of Monopoly And Regulation, Atlantic Economic Journal, 8, 22-26.
BERGSON, A. (1973) On Monopoly Welfare Losses, American Economic
Review, 63, 853-70.
BROWN, D. J. and WOOD, G. A., (2004a) The Social Cost of Monopoly
Power, Yale ICF Working Paper No. 04-31; Cowles Foundation
Discussion Paper No. 1466.
BROWN, D. J. and WOOD, G. A., (2004b) Competition, Consumer
Welfare and Monopoly Power, Cowles Foundation Discussion Paper
No. 1466R.
BROWNING, E.K. (1997) A neglected welfare cost of monopoly-and most
other product market distortions, Journal of Public Economics, 66, 127144.
COWLING, KEITH AND MUELLER D. C. (1978) The Social Costs of
Monopoly Power, The Economic Journal, 88, 727-748.
DAVID S. (1960) The Burden of Monopoly, Journal of Political Economy, 68,
627-30.
DAVID S. (1961) The Effect of Monopoly; A Correction, Journal of Political
Economy, 494.
DAVIS, D. D. AND R. J. REILLY (2000) Multiple Buyers, Rent-Defending
and the Observed Social Cost of Monopoly, Pacific Economic Review,
5(3), 389-410.
DEBREU, G. (1951) The Coefficient of Resource Utilization, Econometrica,
19, 273-292.
DIXON, R., A. GUNTHER, AND M. MAHMOOD (2001) The Cost of
Monopoly in Australian Manufacturing, The Australian Economic Review,
34(4), 390-402.
163

The Social Costs Of Monopoly: A Survey And An Evaluation

EPSTEIN, R. C. (1934) Industrial Profits in the United States, NBER.

GEORGE J. S. (1956) The Statistics of Monopoly and Merger, Journal of
Political Economy, 64(1), 33-40.
HARBERGER, A. C. (1954) Monopoly and Resource Allocation, American
Economic Review, 77-87.
KAMERSCHEN, D. R. (1966) An Estimation of the Welfare Losses from
Monopoly in the American Economy, Western Economic Journal, 221236.
KEEM, J. H. (2001) The Social cost of monopoly when consumers resist,
European journal of political economy, 17, 633-639.
LEE, YOON-HO ALEX and BROWN, DONALD J. (2006) Competition,
Consumer Welfare, and the Social Cost of Monopoly, Yale Law School:
Yale Law School Student Scholarship Series. Paper 19. Available at
http://lsr.nellco.org/yale/student/papers/19 (last visited, 28th of July,
2006).
POSNER, R. A. (1975) The Social Costs of Monopoly and Regulation,
Journal of Political Economy, 83(4), 807-827.
SIEGFRIED, J. J. AND TIEMANN T. K. (1974) The Welfare Cost of
Monopoly: An Inter-Industry Analysis, Economic Inquiry, 12, 190-202.
SHOVEN, J. AND J. WHALLEY (1992), Applying General Equilibrium,
Cambridge University Press.
STEGEMANN, K. (1984) The Social Costs of Monopoly in an Open
Economy, Canadian Journal of Economics, 17, 718-730.
TULLOCK, G. (1967) The Welfare Costs of Tariffs, Monopolies, and
Theft, Western Economic Journal 5, 224-32.
WENDERS, J. T. (1987) On Perfect Rent Dissipation, The American
Economic Review, 77(3), 456-459.
WORCESTER, DEAN A. JR. (1973) New Estimates of the Welfare loss to
Monopoly, United States: 19561969, Southern Economic Journal, 40,
234-45.
YOON, S. (2004) Welfare Losses due to Monopoly: Koreas Case, Journal of
Policy Modeling, 26, 945-957.
YOUNG, D. P.T. (1997) Dominant firms, price leadership and the
measurement of monopoly Welfare losses, International Journal of
Industrial Organization, 15, 533-547.

164