Market Power in
the Australian
Food Chain:
Towards a Research Agenda
A report for the Rural Industries Research
and Development Corporation
by Roley Piggott, Garry Griffith and
John Nightingale
October 2000
RIRDC Publication No 00/150
RIRDC Project No. UNE-67A
© 2000 Rural Industries Research and Development Corporation.
All rights reserved.
ISBN 0 642 58180 0
ISSN 1440-6845
Market Power in the Australian Food Chain: Towards a Research Agenda
Publication No. 00/150
Project No. UNE-67A.
The views expressed and the conclusions reached in this publication are those of the author and not
necessarily those of persons consulted. RIRDC shall not be responsible in any way whatsoever to any person
who relies in whole or in part on the contents of this report.
This publication is copyright. However, RIRDC encourages wide dissemination of its research, providing the
Corporation is clearly acknowledged. For any other enquiries concerning reproduction, contact the
Publications Manager on phone 02 6272 3186.
Researcher Contact Details
Roley Piggott
School of Economic Studies
The University of New England
Armidale NSW 2351
Australia
Phone: 02-67732313
Fax:02 67733596
Email:[email protected]
RIRDC Contact Details
Rural Industries Research and Development Corporation
Level 1, AMA House
42 Macquarie Street
BARTON ACT 2600
PO Box 4776
KINGSTON ACT 2604
Phone:
Fax:
Email:
Web:
02 6272 4539
02 6272 5877
[email protected].
http://www.rirdc.gov.au
Published in October 2000
Printed on environmentally friendly paper by Canprint
ii
Foreword
Whether or not market power exists in the food chain has been an issue of concern to farmers and
policy makers for most of this century The broad aims of this project were (a) to try to determine
whether current knowledge allows a reliable conclusion to be made about the existence and effects of
market power in Australian food markets through undertaking a literature review; (b) to suggest an
agenda for future research on this issue and to present the agenda to a workshop for critical review.
Over time the food marketing chain and how it is managed has undergone significant change and
change continues to occur. Once, food was "pushed" off farms into the marketing chain with
consumers accepting whatever was on offer. Now retailers "pull" product with appropriate
characteristics out of the system in response to consumer preferences.
Questions that are being addressed include whether market power is increasing or decreasing as a
result of changes in the structure and management of the food chain, and how various parties (farmers
and consumers in particular) are affected. Public concern about the issue is evidenced by the Joint
Select Committee on the Retailing Sector which reported in 1999. The structure of food retailing and
the agricultural marketing board system also received attention in the 1999 Productivity Commission
report on the impacts of National Competition Policy on rural and regional Australia.
Interest has continued on other fronts as well. There have been a number of recent reports on the
progress of the German retailer Aldi, in purchasing stores in Sydney, on opening a large distribution
warehouse, and in planning to open 100 stores in NSW metropolitan centres (Western Sydney,
Wollongong and Newcastle) over the next few years, and the impact this might have on retail
competition in the food chain. Also, there is continuing interest in investments in on-line food
shopping.
After reviewing relevant economic theory and empirical work on market power in the food chain,
attention is given to a research agenda aimed at throwing further light on the issues raised.
This project was funded from RIRDC Core Funds which are provided by the Federal Government and
is an addition to RIRDCs diverse range of around 600 research publications. It forms part of our
Global Competitiveness R&D program, which aims to identify important impediments to the
development of a globally competitive Australian agricultural sector and support research that will
lead to options and strategies that will remove these impediments.
Most of our publications are available for viewing, downloading or purchasing online through our
website:
•
•
downloads at www.rirdc.gov.au/reports/Index.htm
purchases at www.rirdc.gov.au/eshop
Peter Core
Managing Director
Rural Industries Research and Development Corporation
iii
Acknowledgments
The authors wish to acknowledge the financial assistance of the Rural Industries Research and
Development Corporation and the support and constructive comments of Dr Jeff Davis, General
Manager, Research.
The project team was assisted by an Industry Advisory Committee comprising Rhonda Smith,
Australian Competition and Consumer Commission; Jim Booth, NSW Cabinet Office; Todd Richie,
National Farmers Federation; and Michael O’Keefe, Franklins Foods. These people gave freely of
their time and experience and made substantial contributions to the project, through provision of
articles and reports, attendance at meetings and the Workshop, and reading and commenting on draft
sections of the report.
Attendees at the Industry Workshop provided a number of different perspectives on the literature
reviews, the issues raised and the proposed research agenda, and they are all acknowledged for their
contributions. Their names are listed in an Appendix. Michael Keogh and his staff at NSW Farmers’
Association generously provided the venue for the Workshop.
Adam Turcato and Quang Ng provided enthusiastic and valuable assistance with the literature search
and collating tasks.
Abbreviations
ABARE
Australian Bureau of Agricultural and Resource Economics
ABS
Australian Bureau of Statistics
ACCC
Australian Competition and Consumer Commission
EBIT
Earnings Before Income Tax
ECR
Efficient Consumer Response
NARGA
National Association of Retail Grocers Australia
NFF
National Farmers' Federation
NEIO
New Empirical Industrial Organisation
OECD
Organisation for Economic Co-operation and Development
QFVG
Queensland Fruit and Vegetable Growers
R&D
Research and Development
SCP
Structure-Conduct-Performance
SUR
Seemingly Unrelated Regressions
UK
United Kingdom
US
United States
iv
Contents
Foreword ........................................................................................................................................... iii
Acknowledgments ..............................................................................................................................iv
Abbreviations .....................................................................................................................................iv
List of Tables......................................................................................................................................vi
Executive Summary ..........................................................................................................................vii
1: Introduction ................................................................................................................................... 1
2: Insights from Economic Theory ..................................................................................................... 3
2.2 Elementary Theory .............................................................................................................. 5
2.3 Models of Vertically-Related Markets ................................................................................... 5
2.3.1 Location Factors ........................................................................................................ 7
2.3.2 Size Distribution of Firms ........................................................................................... 7
2.3.3 Household Production ............................................................................................... 7
2.4 The Structure–Conduct–Performance (SCP) Framework ..................................................... 8
2.5 Contestability Theory..........................................................................................................10
2.6 New Institutional Economics ...............................................................................................12
2.7 Evolutionary Economics .....................................................................................................13
2.8 Policy Implications..............................................................................................................15
3: Insights from Empirical Studies.....................................................................................................17
3.1 Traditional Structure-Conduct-Performance Studies............................................................17
3.2 New Empirical Industrial Organisation Studies ....................................................................18
3.2.1 Examples of US and Canadian Studies.................................................................... 19
3.2.2 Examples of European Studies ................................................................................ 21
3.2.3 Examples of Australian Studies................................................................................ 21
3.3 Time Series Methods..........................................................................................................21
3.4 Conclusions........................................................................................................................22
4: Recent Australian Inquiries...........................................................................................................25
4.1 The Joint Select Committee on the Retailing Sector............................................................25
4.1.1 Winners and Losers................................................................................................. 25
4.1.2 Methods .................................................................................................................. 26
4.1.3 ACCC Submission................................................................................................... 26
4.1.4 Recommendations................................................................................................... 26
4.1.5 Other Comments ..................................................................................................... 27
4.2 Review of National Competition Policy (NCP) on Rural/Regional Australia ..........................28
5: The Research Agenda..................................................................................................................29
5.1 How Does One Confirm the Existence of Market Power?....................................................29
5.1.1 Is There Evidence of Long-Run Economic Profits?................................................... 29
5.1.2 Profit Sources.......................................................................................................... 29
5.1.3 Profit Nodes............................................................................................................. 29
5.2 What Causes Market Power? .............................................................................................30
5.2.1 Barriers to Entry to Retailing and Processing ........................................................... 30
5.2.2 Importance of Scale of Operation............................................................................. 30
5.2.3 What are the Market Niches in the IT Age? .............................................................. 30
5.2.4 Can Alternative Retailing Systems Co-exist?............................................................ 30
5.2.5 Do House Brands Matter? ....................................................................................... 30
5.2.6 Do Regulations Matter? ........................................................................................... 30
5.2.7 Demand Elasticity.................................................................................................... 31
5.3 Is Contracting an Effective Response to Market Power? .....................................................31
5.3.1 Case Studies ........................................................................................................... 31
5.3.2 Contracts................................................................................................................. 31
5.3.3 Alternative Institutional Relationships....................................................................... 31
5.3.4 Supply Elasticity ...................................................................................................... 31
5.4 Views at the Workshop .......................................................................................................32
v
6. Appendices ..................................................................................................................................33
6.1 Appendix A: Changes in Food-Chain Management Around the World .................................33
6.2 Appendix B: A Model of Marketing Margins Under Perfectly Competitive and Pure
Monopoly/Monopsony Behaviour..................................................................................................37
6.3 Appendix C: Some Comparative Statics on the Incentive for Collusive Buying Behaviour...47
6.4 Appendix D: Some New Empirical Results.........................................................................49
6.5 Appendix E: The Workshop ...............................................................................................55
7. References...................................................................................................................................56
List of Tables
Table B.1:
Table B.2:
Table B.3:
Table B.4:
Table D1:
Total Elasticity Formulae ................................................................................................. 42
Implied Values for β 1 and β 2 ........................................................................................ 43
Base Solutions................................................................................................................ 43
Total Elasticities.............................................................................................................. 44
Constrained SUR Estimates............................................................................................ 52
vi
Executive Summary
In the main body of the report, insights about market power gleaned from economic theory and
empirical studies on market power in the food chain are reviewed in turn. Recent Australian public
inquiries having some coverage of food chain market power are then discussed. Finally, a research
agenda is presented along with an overview of discussion on the agenda that took place at a workshop
convened for that purpose.
The body of the report is supported by five appendices. Changes in food chain management are
reviewed. A model of marketing margin behaviour under extremes of competition in retailing and
input purchasing and some comparative statics relating to the incentive for collusive behaviour in
purchasing raw agricultural products are presented. Some preliminary empirical work on market
power in the Australian food chain is presented. Finally, some details of the workshop convened to
review the research agenda are provided.
Future research into market power issues in the Australian food market should be in the form of
intensive case studies. These will allow development of the institutional detail sufficient to reveal the
nature and significance of market power in those chains and how and why it ebbs and flows over time.
However, such resource-hungry analyses cannot be undertaken for every market faced by a merger or
deregulation proposal.
The implications of the empirical results provided in the Report are that (a) attention should be
focused on the supplier side, not on the consumer side of the market (an implication fully supported in
the Joint Select Committee report); and (b) the sector providing the most likely payoffs from greater
research effort would be the grains and oilseeds processing and distribution sector. Lack of empirical
data for many other sectors means that we depend on anecdote and intuition for suggestions that
horticulture might be another significant sector.
vii
1: Introduction
Whether or not market power exists in the food chain has been an issue of concern to farmers and
policy makers for most of this century. Indeed, the incentive for the establishment of marketing
boards for primary products was partly the fear that farmers were at the mercy of powerful players in
the food marketing chain who had the ability to earn abnormally high profits at the expense of
farmers.
Over time the food marketing chain and how it is managed has undergone significant change and
change continues to occur. Once it was common to describe the food marketing chain as
encompassing all of the activities involved in moving raw farm products from the "farm gate" to the
"consumer's kitchen". But this description is growing increasingly inappropriate. For one thing it
does not acknowledge the fact that food marketing firms, sometimes acting in response to consumer
preferences and sometimes of their own volition, play an important role in determining what is
actually produced on farms. For another, it ignores the trends, such as eating food away from home
and concerns for food safety, which are manifestations of income growth, time-constrained
households and more discerning consumers. Finally, it creates the false impression that food
marketing is much like a motor vehicle production line: activities start at one end and follow a set
pattern through until the finished product is available for consumers. The food marketing chain today
comprises a closely-integrated network of vertically-related players extending from the farm to the
retail outlet. There is significant concentration at the retail and processor levels, and a complex web of
channels, some of which bypass most intermediate levels in the chain, while others pass through many
more levels than simply farmer-processor-wholesaler-retailer.
Paraphrasing Kinsey (1996), it was once the case that food was "pushed" off farms into the marketing
chain with consumers accepting whatever was on offer, often at discounted prices that cleared the
market. It is now more a case of retailers "pulling" product with appropriate characteristics out of the
system in response to consumer preferences. An overview of changes taking place in the food
marketing chain is provided in Appendix A.
Concern about market power in the food chain has not diminished despite the changes that are
occurring. In the US, for example, there are various research centres focussing on this issue and it is
an item of on-going concern for various government agencies (see, for example, US Department of
Agriculture 1996, 1999a, 1999b). Questions that are being addressed include whether market power is
increasing or decreasing as a result of changes in the structure and management of the food chain, and
how various parties (farmers and consumers in particular) are affected. Whilst market power in the
food chain has not received quite the same degree of attention among economists in Australia, there is
certainly public concern about the issue as evidenced by the Joint Select Committee on the Retailing
Sector which reported recently (see Australian Parliament 1999). The structure of food retailing and
the agricultural marketing board system also received attention in the recently-completed Productivity
Commission report on the impacts of National Competition Policy on rural and regional Australia
(Productivity Commission 1999).
Recent press reports highlight the continuing concern in the community about the possiblity of the
abuse of market power in the Australian food chain. For example, in The Land newspaper of 8 June
2000, there were reports of the NSW Dairy Farmers Association requesting the ACCC to become
involved in negotiations of farm gate milk prices following deregulation to provide countervailing
power against the supermarkets. In this report it was reported that the ACCC “conceded that farmers
were being subjected to exceptional price bargaining pressures from the supermarkets.” In the same
issue, there was a report of another fruitgrower meeting raising the issue of the large differences
between farm and retail apple prices and the apparent “excessive” supermarket markups on fresh fruit.
As argued in our report, oligopsony power by processors and retailers is seen as the major issue in the
1
food marketing chain. A number of reports have also pointed to the crucial role of the Retail
Ombudsman recommended by the Joint Select Committee on the Retailing Sector.
Interest has continued on other fronts as well. There have been a number of recent reports on the
progress of the German retailer Aldi, in purchasing stores in Sydney, on opening a large distribution
warehouse, and in planning to open 100 stores in NSW metropolitan centres (Western Sydney,
Wollongong and Newcastle) over the next few years, and the impact this might have on retail
competition in the food chain. Also, there is continuing interest in investments in on-line food
shopping.
It is against this background that the authors decided to: (a) review the economics literature on market
power, or at least that part of it thought to be most relevant to the food chain; (b) map out a research
agenda relating to the economic analysis of market power in the Australian food chain; and (c) present
the results to a workshop attended by a broad cross-section of individuals with interests in the general
topic of market power in the Australian food chain. The motivation for this work was partly because
of the changes that have occurred in food chain structure and management, partly because the area has
not attracted the same degree of attention as it has in some other western countries and partly because
of changes that are occurring in the way of dismantling the agricultural marketing board system. The
latter has been happening over a period of about ten years and has been spurred on by National
Competition Policy. It does seem that farmers will be forced to consider marketing issues much more
intensively than they have had to under the marketing board system and the changes that are occurring
in food chain structure and management will make such consideration even more imperative.
The remainder of the report is structured as follows. In Chapters 2 insights about market power
gleaned from economic theory are summarised. In Chapter 3 empirical studies on market power in the
food chain are reviewed. Recent Australian public inquiries having some coverage of food chain
market power are discussed in Chapter 4. A research agenda is presented in Chapter 5. The body of
the report is supported by five appendices. Changes in food chain management are reviewed in
Appendix A. A model of marketing margin behaviour under extremes of competition in retailing and
input purchasing is presented in Appendix B. Some comparative statics relating to the incentive for
collusive behaviour in purchasing raw agricultural products is presented in Appendix C. Some
preliminary empirical work on market power in the Australian food chain is presented in Appendix D.
Finally, some details of the workshop convened to review the research agenda mapped out in Chapter
5 are presented in Appendix E.
2
2: Insights from Economic Theory
This chapter begins with a definition of "market power" and then identifies what various strands of
economic theory have to offer on the subject. In principle, at least a couple of centuries of writings by
economists about market power could be reviewed. Moreover, the definitive word on the subject is
yet to appear, and may never appear. Nevertheless, there have been many useful insights. They are
summarised here. Because of the enormous literature, the focus is on summarising ideas from the
various strands; not much attention is given to particular studies.
2.1
Defining Market Power
The definition of market power is never clear in economic theory. One definition that has been used
by the ACCC (1999) is "…the ability of a firm to behave persistently in a manner different from the
behaviour that a competitive market would enforce on a corporation facing otherwise similar cost and
demand conditions." This would include the ability to raise selling prices and depress input prices, to
deter entry, to re-distribute profits to oneself from other players, and, importantly, to be able to sustain
these benefits over time. This last point is most important. If a market player takes advantage of a
temporary situation of power, it will have less effect on the well-being of other buyers and/or sellers
than would a permanent advantage. It could be said that market plays are a continual effort to make
the best of the present situation. Only if the advantage stays with one player, or set of players, is the
market frustrated in allocating resources efficiently. A few points are made here to emphasise the
diversity of ways in which market power may be exercised.
The most obvious way market power is exercised is through elevating the price of a product above that
which is required to encourage the firm to produce the product, given input prices. Elementary
economics textbooks illustrate this in their analysis of monopoly. Similarly, a firm which is able to
exert monopsony power can depress the price of monopsonised inputs below the level which a
competitive market purchaser would pay when buying from a competitive supplying market. These
simple cases ignore the historical time period during which such behaviour may take place. The first
elaboration of them is therefore to suggest that if these behaviours are sustainable over a considerable
period they may constitute a more serious loss to stakeholders and society, and unnecessary transfers
of profits to the power holder, than if they are wiped out quickly by market forces. Understanding
how long it may be before the market takes its revenge on short run power is by no means simple.
A second common use of market power is to discourage competitors from entering the market, or
encouraging them to leave. Unfortunately, this behaviour is often identical in form to the process of
competition, whereby firms attempt to produce products that will dominate those of competitors, and
also convince potential suppliers that they have no adequate way of producing a credible substitute.
Firms which invest in new techniques of production or marketing, or which advertise their products
and services so successfully that they create a product differentiation barrier to entry, are merely
successful competitors in an oligopolistic market. If they are so successful that their long-run profits
are protected, should regulators try to do what competitors have failed to do? On the other hand,
behaviour that is designed to exclude competition is clearly socially undesirable. A firm which
corners supplies of essential inputs, or abuses intellectual property law to prevent use of innovative
technologies (patents are not always taken out simply to ensure a fair return on investment in
invention), or enters into exclusive arrangements for distribution, is clearly abusing market power.
Yet another form of competition which can be used to bolster market power is an emphasis on market
share by a dominant firm, where market share is purchased in the short term by pricing and marketing
policies which in themselves appear extremely rivalrous. It is clearly very difficult to accuse a
successful firm of monopolising its market by selling at too low a price! It may be a little easier to see
how one firm’s proliferation of brands may foreclose a market to competing firms. Filling all the
niches, thus forcing competitors to attack head on, whichever niche they choose, can be seen as an
option open only to firms with considerable power derived from size and financial strength. But even
3
here, whether it is really detrimental to consumers depends on the long-run sustainability of the
strategy.
The Microsoft case in the US illustrates some of the problems involved in determining whether market
power is being exercised. Microsoft is accused of forcing personal computer manufacturers to use its
web browser software, of bundling that software with its system software, and of threatening related
firms if they did not agree to Microsoft conditions of supply. Plaintiffs and witnesses at the trial have
cited numerous other instances of apparent abuse of market power. Microsoft appears to absorb
potential competition by adopting and then modifying previously standard systems to make interworking more difficult. Examples include Java and Kerberos software (The Australian, 4 April 2000,
p. 55; 5 April 2000, p. 32). The context is that Microsoft is the de facto monopolist for system
software for nearly 90 per cent of personal computers. The reason is that common software facilitates
networking of computers. These network economies, therefore, imply that this dominance is a natural
outcome of market forces. Moreover, Microsoft argues that they did not set out to be dominant, and
that it flows mainly from rapid innovation in system software and ancillary areas such as Local Area
Network (LAN) server software. If they are not allowed to enjoy the dominance the market has
delivered, they argue that the rate of innovation will fall away, network economies will be sacrificed,
and productivity growth in the global economy will be reduced. Opinions differ as to the inevitability
of the virtual monopoly of the Bell system in US telephony (Evans 1983), but network economies
were the feature of the system which encouraged AT&T to pursue the monopoly it enjoyed for most of
the twentieth century.
The ability of a dominant firm to lock-in its own technology is an important element in a system
embodying network economies. This may happen whether the eventually dominant firm is trying to
exercise dominance or not. If a potential participant has to access the network, it must conform to the
requirements of that network. The ability of the dominant firm to extract a rental payment for access
to the technology, whether explicit or implicit, marks an abuse of market power. Here intellectual
property law may in some cases be an unwitting accomplice to the exercise of power. In other cases
the power comes from more economic factors, such as the participants having to invest in additional
technology to satisfy network requirements, consequent on the dominant firm’s early choices.
Some aspects of "Just In Time Delivery" and other logistical systems used by major players in the
food chain may have this characteristic for small food processors and distributors. Retail chain
requirements for delivery frequency, time of day, minimum volumes (pallet numbers), restrictions on
what items may be delivered together (pallet contents), on delivering to different departments, and on
the type of delivery packaging, all potentially impose costs on suppliers. The smaller the supplier, the
greater the burden of any fixed cost of compliance. The entry of Aldi to the Sydney grocery market
(see Appendix A) may lead to some potential suppliers deciding not to try to supply. The Aldi method
for at least part of its product range is a whole pallet delivered direct to the customer-accessed aisle.
Dimensions and construction of the pallet may be distinctive, and the cost shifted to suppliers.
It is quite understandable that dominant firms would choose to invest in techniques of production that
enhance their position; for example, techniques which enhance economies of scale or scope. Some of
the above instances represent this phenomenon. In the "normal" case, in the absence of network
economies or close co-ordination, investment in economies of scale and scope raise the price of entry,
and ensure that actual entrants have to be well capitalised in order to invest in cost-efficient
technology. In the network, or close co-ordination, case the effect is stronger. Potential entrants must
not only invest in cost-efficient production technology but also invest in additional technology
required for the network.
The Microsoft case also illustrates the dilemma faced by social policy makers. The company clearly
has enormous market power, but it argues that that power is an unintended consequence of its
successful policies of rapid innovation. These policies have resulted in a long-term downward spiral
of costs and prices for its outputs and those of the firms within the sector which defines the network of
4
interlocked economic activity dominated by Microsoft. Judge Jackson has made his judgement, that
Microsoft has violated US federal anti-trust law, the 1890 Sherman Act.
But the problem remains, how should market power be recognised and how does society determine
what to do about it? Economic theory and empirical work based on it does give us some basis for
action. The remainder of this chapter attempts to describe the insights theory can provide. The
following chapter reviews the empirical work based on some of these ideas.
2.2 Elementary Theory
The concept of monopoly in introductory economics textbooks has important implications for the
assessment of market power in the food marketing chain. The implication of profit maximisation
under monopoly is that the monopolist will hold price higher than justified by costs, and produce an
output smaller than would otherwise be produced under competitive conditions. Authors such as
Edwin Chamberlin (Chamberlin 1933) and Joan Robinson (Robinson 1933) refined the concept of the
losses suffered from this form of monopoly power, suggesting means for measurement of this loss.
Arnold Harberger’s analysis of the deadweight loss triangle for the US economy (Harberger 1954)
was the first of many attempts at measurement. He showed that the loss is rather trivial, of the order
of one to five per cent of GDP if the whole economy was comprised of monopolists. Many variations
on Harberger’s measurement followed, searching for something more substantial. Scherer and Ross
(1990, pp. 661-7) review this literature on the welfare costs of monopoly power under this essentially
static view of a market in equilibrium. The conclusion remains much the same as that of Harberger:
the measured costs of monopoly under this definition of monopoly cost are not very important.
Nonetheless, for a mature and stable industry exhibiting high levels of market concentration, these
losses may be important and the deadweight loss triangle a valid measure of market power.
The use of market concentration as a measure of market power is suggested by this simple view of
markets. If a market is monopolised, the absolute price distortion caused by profit-maximising
behaviour is price divided by the price elasticity of demand. Alternatively, one can say that the price
elasticity of demand is inversely related to the price-cost margin. For products whose price elasticity
of demand is low, the price distortion is expected to be larger than for highly price-elastic products. In
the present context, food commodities broadly defined will have a relatively low price elasticity, while
food product lines will have highly-elastic demands. Monopolisation, or cartelisation of a
concentrated food retailing industry, will therefore be important, as this industry determines the final
level of prices of the broad range of food it retails. This piece of elementary economics may be one
factor driving politically-important consumers to have an interest in the structure of food processing
and retailing.
2.3 Models of Vertically-Related Markets
The neoclassical analysis of vertically-related markets provides a first base from which further work
on these markets may be done. The approach is a structural one: by looking at the structure of the
vertically-related market, and making the usual assumptions about perfect knowledge and exogeneity
of the institutions of the market, the implications for prices and outputs, given profit maximisation,
can be made. The pernicious results of vertically interacting market power are a powerful argument
against unthinking opposition to vertical integration: If both an upstream production stage and a
downstream stage are monopolised, the effect of setting marginal revenue equal to marginal cost when
marginal revenue is less than price at both stages is to compound the monopoly restriction on output
and elevation of price. The first stage monopolist produces less at a higher price than would a
competitor. If that monopolist’s market includes a firm which is a monopolist in its own market (the
first stage monopolist’s output is an input to the second stage) then that second stage monopolist will
use less of the over-priced input than otherwise. It will then sell its own output at an elevated price
reflecting both the over-priced input and its own market power. The deadweight loss from the first
stage, upstream, monopolist’s restriction will be compounded by the restriction and input choice bias
5
of the second stage, downstream, monopolist. An example would be a monopoly marketing board
selling its output to various processors, amongst which is one which is a monopolist in its regional
market. This compounding of the deadweight loss can have a significant effect on final price and
output. Vertical integration removes the upstream distortion, as the integrated firm maximises its
surplus in producing both the first and second stage outputs.
Even more of a problem is bilateral monopoly, where both monopoly seller and monopsony buyer of
an input are inclined to minimise the output they offer, and the bargain is about distributing the surplus
by a lower or higher price. Vertical integration internalises the surplus and removes the incentive to
restrict quantity offered or accepted. While some aspects of market power may be enhanced by such
integration, the benefits in terms of eliminating these incentives to take decisions on marginal
profitability calculus will need to be discounted before such vertical integration is disallowed by social
policy.
The most common way that agricultural economists have addressed vertical market relationships and
market power in the food chain is through the use of so-called marketing margin analysis. The
marketing margin can be defined as the difference in the price of a food item at two different market
levels (e.g., wholesale and retail prices for flour). The size of marketing margins is an on-going
concern to farmers.
In general the theory of marketing margins is better developed for extreme cases of perfect
competition and pure monopoly and/or monopsony. The theory of marketing margins when there are
departures from price-taking behaviour, such as when a retailer has some monopoly power or a food
processing firm has some monopsony power, is less-well developed since much depends on how rival
firms react to each others pricing, purchase or output decisions. Different assumptions concerning
these reactions lead to different predictions regarding market outcomes.
Concern about the size of marketing margins (i.e., that they are too high) was part of the driving force
behind two major US studies on the competitiveness of food marketing. The first was the National
Commission on Food Marketing which reported in 1966 (National Commission on Food Marketing
1996), producing a main report and eight supporting technical papers. It found the food marketing
system in the US to be "workably competitive" but had concerns about the amount of advertising and
product differentiation, and trends towards centralisation of manufacturing and distribution (see
Connor et al. 1985, pp. xvii). The second major study commenced in 1973 under the auspices of the
(US) North Central Regional Research Project (NC 117) and resulted in a mass of publications. Some
leading contributors to this project summarised many of the findings in book form (Connor et al.
1985). The book highlighted the mergers that began to take place soon after the National Commission
reported and the conglomerate nature of the big players in food manufacturing and distribution.
Evidence of market power (eg, in the procurement of horticultural products) was documented.
It is probably fair to say that many unwarranted conclusions about (a) the behaviour of food marketing
firms and (b) the welfare of farmers have resulted from misunderstandings of the nature of marketing
margins. Farmers and their industry bodies have generally been less than careful in the way they have
interpreted increased margins as an indicator of the exploitation of market power at the expense of
farmers. Sometimes, for example, an increased marketing margin may simply reflect an increase in
demand for more marketing services. In this situation, farm-level prices do not change although there
would likely be an increase in retail prices and, hence, a greater marketing margin.
Tomek and Robinson (1990, Chapter 6) provide an excellent introduction to the topic of marketing
margins under conditions of perfect competition. They also demonstrate the dangers involved in the
use of concepts such as "the farmer's share of the consumer dollar". One of the important issues in
marketing margin analysis concerns the impact or incidence of changes in marketing margins on
prices at various levels. For example, if a more efficient method of peeling and coring pears was
invented, how would this affect farm–level and retail–level prices? Tomek and Robinson provide a
6
graphical analysis of the incidence of changes in margins, although this needs to be supplemented by
the paper by Fisher (1981).
While the analysis provided by Tomek and Robinson remains perhaps the clearest available at an
introductory level, it is important to appreciate a crucial assumption that they employ. The assumption
is that there is zero substitution between the raw agricultural product and other inputs in producing the
final product demanded by consumers. In technical terms, the elasticity of substitution between inputs
is zero. This makes for convenient graphical analysis as in Tomek and Robinson (Chapter 6) since the
farm–level demand function can be derived directly from the final or primary demand function by
subtracting the cost of marketing services. Likewise, the derived supply of the final product can be
derived from the primary supply of the farm product by adding the cost of marketing services.
Is substitution common? Tomek and Robinson believe that the assumption of zero substitution is
fairly realistic (see their footnote on p.108), although they acknowledge that a high price for the farm
product may provide an incentive to use more labour to prevent wastage and spoilage. For example, a
beef carcass might be trimmed more thoroughly using extra labour if livestock prices are high: this is
substitution of labour for the farm product.
Wohlgenant and Haidaicher (1989) addressed this general question in a US Department of Agriculture
technical bulletin and the thrust of their comments were that agricultural economists have been rather
narrow–minded in their thinking about the extent of substitution between farm and non–farm inputs in
producing retail products. Whilst acknowledging efforts to reduce wastage and spoilage as a clear
example of substitution, they mentioned a number of other sources.
2.3.1 Location Factors
Farmers located at different points relative to markets might use different transport methods which
amount to different combinations of the farm product and marketing services. For example, cutflower producers in Tasmania air freight their product to Melbourne markets while Victorian
producers use road transport. The Tasmanian producers can, at least to some extent, use sea transport
if the price of air freight increased sufficiently.
2.3.2 Size Distribution of Firms
Even if individual firms of a given size use inputs in fixed proportions, firms of different sizes may
use a different combination of inputs. Thus, one would be likely to observe substitution at the industry
level as relative prices change and firms of certain size groups enter or leave an industry. An example
would be small-scale vs large-scale fruit producers; different production technologies are used in each
(particularly, the capital/labour ratio is different). Over time there has been an exodus of small scale
producers and so, at the industry level, there has been substitution of capital for labour.
2.3.3 Household Production
Substitution of the raw farm product for marketing inputs can result indirectly from the phenomenon
of household production of food services. Less food preparation is done at home, and more food
marketing services are purchased, as incomes grow and, hence, ought to be an important source of
substitution in developing countries.
In short, Wohlgenant and Haidacher argue that substitution between the farm product and other
marketing inputs is more widespread than has been commonly thought.
A publication by Gardner (1975) was a turning point in the analysis of marketing margins. Gardner
drew attention to the limitations of the Tomek and Robinson-type analysis and presented a model of
marketing margin behaviour in a competitive marketing system which allowed for substitutability
between inputs (the raw product produced by farmers and other marketing inputs) in producing the
7
final product. It is important to understand that the type of analysis provided in texts such as Tomek
and Robinson really covers a "special case" of marketing margin behaviour (the fixed proportions
case). Gardner's model of marketing margins builds on earlier models of a competitive industry to
analyse the connection between supply elasticity and the demand for factors of production (see, in
particular, Muth 1965).
In Appendix B a model of marketing margin behaviour is set out that contains the original Gardner
model as a special case. The more general model accommodates extremes of competition in the
market for the final product and the market for the farm-produced input used in making the final
product. The model development draws on suggestions made in the original Gardner paper on how
his analysis could be extended to cover monopoly in the sale of the final product and monopsony in
the purchase of the farm-produced input. The Appendix includes the case of perfect competition in
final product sale and input purchase (Gardner's case) and this is the scenario that most commentators
regard as providing the benchmark against which results of other scenarios can be compared. Another
scenario that is included is monopoly in sale of the final product and monopsony in the purchase of the
farm-produced input and this is, presumably, the scenario most feared by consumers and farmers.
2.4 The Structure–Conduct–Performance (SCP) Framework
In the US, concerns about monopoly power led to the anti-trust movement of the last century. The
Sherman Act, and those following, created a body of law which had loose connections with economics
until E. S. Mason (1939, 1949) introduced the structure-conduct-performance framework. Under this
schema, the underlying structure of an industry was determined by technological and physical factors.
Scale of production was seen as technically determined. The structure of an industry therefore
included the size of firms, the number of firms that could serve the market and the nature of the
products (homogeneous or heterogeneous). Given structure, conduct was implied. A monopoly
industry would be expected to behave as such, but with a large number of similar firms, competitive
behaviour could be seen. Given conduct, performance outcomes were seen. Scherer and Ross (1990,
p. 5) present a diagram outlining the framework.
This was elaborated by many authors, including Joe S. Bain, whose Barriers to New Competition
(1956) examined the sources of market power in terms of its sustainability. Barriers to entry by new
firms constituted the protection enjoyed by incumbents against competitive forces. Where barriers
were very high, incumbent firms could ignore entry threats and behave as a cartel much more easily
than if new firm entry was threatening. The most interesting result of Bain’s work was the clear
importance of product differentiation as a barrier to entry: the barrier posed by consumer acceptance
of a well-known brand, allowing a premium on price above that which an unbranded product or less
well-known product could gain.
This framework continues to be elaborated and modified, as shown by Scherer and Ross (1990), the
leading textbook in the field. They point out that the causation runs not only from structure to
conduct, but in the reverse direction, as technological and institutional change is by no means
exogenous. It is worth examining their Chapter 17 in some detail, as it is clear that if market power is
to be significant in the long run, it will be due to its effect on innovation and its diffusion, in turn
lowering factor productivity and real incomes permanently.
A market in which significant players use their profits for R&D more than for capacity expansion or
diversification will develop new structures as R&D reveals alternative growth paths. The concept of
the patent race is of relevance here: where R&D leads to the creation of profitable intellectual
property, the advantage to the patent owner may be enough to grant a permanent increase of market
share or product differentiation barrier. The relationship between patents and market power is a close
one, as the patent right is a monopoly right. Scherer and Ross point out (pp. 624-6) that US
experience has been that cross-licensing of patents has been abused for purposes of exclusion, and that
US antitrust law has "dealt harshly" with this form of abuse. Whether remedies are available in
Australian trade practices law is unclear. A number of provisions of Part IV of the Act may apply, for
8
example, the primary boycott provision or the provision against misuse of market power. In some
circumstances such conduct could be "unconscionable" (Part IVA). The patent system is an historic
part of capitalist economies, but the evidence is that it is necessary to the rate of innovation (pp. 62830). Further evidence that it is abused adds injury to insult. The analysis leading to these conclusions
varies in analytical content from the simple empirical to carefully crafted neoclassical models.
Perhaps of greater importance is the analysis of the relationship between market structure and
innovation (pp. 630-6). Unfortunately, the conclusions of this SCP analysis are not robust. Too many
side conditions affect the judgement about whether more or less highly concentrated industries lead to
innovation being too slow, too fast or just right. The major conclusions are that the speed of
innovation depends on concentration, with monopoly and unconcentrated markets being slow. Some
intermediate level will innovate more quickly than either extreme. Monopoly does not need to rush,
as there is no threat from a rival. Where the industry is concentrated enough for an implicitly cooperative price-output outcome, and returns are high enough to support innovation costs, then
oligopoly will be faster, as the firm’s profits depend on getting a new product to market before its
rivals. Innovation will fail if concentration levels are low enough that the high price equilibrium
breaks down and if returns then become insufficient to support innovation investment.
Another area of significance to the exercise of market power is the relationship between product
differentiation, product innovation and profitability. Chapter 16 of Scherer and Ross explores the
significance of product differentiation from many angles, most of which are significant to concern
about market power in the food marketing chain. The persistence of long run profits has been
identified with product differentiation more than any other phenomenon, though this is by no means a
complete explanation of such profits. First mover advantages (perhaps better called "first successful
mover") appear to come from the habits formed by buyers. These habits apparently persist despite the
lack of objective measures of quality of the goods in question (double blind trials indicate consumers’
inability to distinguish between brands of many consumer goods subject to these advantages) (p. 582).
These first mover advantages appear to be of continuing significance, as evidenced by brands such as
Coca Cola whose advantage has been sustained for something like 80 years in the US market. The
practice of retailers stocking only a very limited range of major brands of any line of product is both a
consequence and a cause of the persistence of this phenomenon.
Galbraith’s concept of countervailing power also came out of the SCP framework. His argument was
that if one side of a market became concentrated, due to structural features or a tendency to invest in
scale-increasing R&D, then the other side of that market would tend to retaliate in kind, devising
institutions for power matching. This has long been thought to be a feature of the food distribution
trades, where initially scale economies in production and downstream vertical integration by
processors created concentrated power over the retail sector. But the retail sector developed
technologies which also advantaged large scale enterprise, and upstream vertical integration. At the
other end of the chain, farmers developed co-operatives to counter the market power of the
concentrated market for the outputs. They did this as a conscious response to market power, with
politically driven assistance from sympathetic governments.
The New Empirical Industrial Organisation economics, with its empirical focus (see Chapter 3 below)
develops oligopoly theory. In the 1960s the theory of oligopoly was developed in directions
unthought of earlier. While it remained the case that theory gave no determined outcome in a given
situation, the number of situations in which bounds could be placed on what might be likely to happen
did increase. Theory, and heavily-analytical models of special situations, was at the centre of concern.
This is not the place for an exposition of that theory. Suffice to say that a game theory approach in
which scale economies, elasticity of demand, price-cost margins, and credibility of threats by one
player against another were all adduced to show how market power could be examined, if not
precisely predicted, in a variety of situations.
There has been some research on the incentive for collusion in the purchase of raw agricultural
products that is very much in the structure-conduct-performance mould. There is clearly much
9
concern among farmers that increased concentration among food marketing firms, especially retailers,
creates an incentive for collusive buying behaviour. Indeed one argument advanced by proponents of
single-desk selling is that such an arrangement is necessary to offset the market power wielded by
concentrated buyers. Central to the argument is the association between concentration and collusive
buying behaviour. However, while collusive buying behaviour might be easier when there is a high
degree of concentration among buyers, the incentive for collusive buying may be lacking. Given that
collusive behaviour entails some costs to those who participate in it, it is unlikely to be undertaken
when the potential gains from doing so are slight.
One would expect the incentive for collusive buying behaviour to be directly related to the potential
impact on price of an expansion in demand. That impact, in turn, is inversely related to the price
elasticity of the supply of the commodity being purchased. Hence, the price elasticity of supply is a
relevant consideration in assessing the incentive toward collusive buying behaviour. This point was
emphasised in an early study which had to do with the potential gains from co-operative bargaining
(see Helmberger and Hoos 1965, pp 129-30 and Piggott 1970):
Greater elasticity in the supply function tends to decrease the difference between average
and marginal resource cost and facilitates independent conduct in the sense that output
variation on the part of any one firm will tend to have a correspondingly smaller impact on
price.
However, price elasticities of demand also play a role in determining the price increase resulting from
an expansion in demand. Too, if the amount of price increase that accompanies an expansion in
demand is of concern to buyers, the price increase that accompanies a contraction in supply would also
be a concern. The relationships involved are explored in Appendix C.
2.5 Contestability Theory
William J. Baumol and his colleagues used a development of some of the ideas from the SCP
framework to explore the consequences of mixing economies of scale with economies of scope, for
the multi-product firm’s ability to credibly deter entry (Baumol, Panzar and Willig, 1982). While the
early contributions to this branch of theory were quite complex, the central ideas are now seen to be
fairly simple, and much less startling than Baumol and his colleagues at the Bell Laboratories had
thought. Sunk costs are the fundamental barrier to entry, considered most broadly; even natural
monopoly may be contested if the source of the economies of scale can be isolated and quarantined by
measures such as dividing an activity into a regulated monopoly subject to economies of scale, and a
contested operation using the regulated infrastructure. In Australia this has resulted in the National
Competition Policy initiative of essential facilities access regulation. However, interest here is not in
this aspect of the contestability controversy, but in the attention it focuses on the assessment of the
importance of sunk costs in protecting market positions. A further element in Baumol’s work on
contestability and the sustainability of a market position is his revival of interest in price
discrimination as a tool of competitive rivalry. Ramsey pricing, as it is known, is the setting of prices
in each relevant market on the basis of the market power of the seller, that is, pricing such that
marginal revenue in each market is equal to marginal cost. This allows some markets to contribute to
covering sunk costs of the business as a whole and, in the extreme, allows average cost to be covered
where uniform pricing would result in a loss. In the present context, where geographically distinct
markets are served by a firm, the rule suggests how a firm may profitably contest a marginally viable
market.
This has led to a long running controversy about when a firm’s behaviour is an exercise of power, and
when it is merely a competitive strategy which the rational firm will employ to produce at minimum
cost for the bundle of products being produced. The idea is, on the surface, attractive, as it suggests
that even a monopolist may have no market power, under certain circumstances. Those circumstances
are where deviation from cost and price minimisation will lead to entry. The form of entry is usually
known as "hit & run". It is where the entrant has no sunk costs. Stated in this way, it suggests that the
10
theory has little application. However, sunk costs may well be quite small in many situations where,
for example, a retailer leases all premises and many fittings, or where an airline leases its fleet and its
airport slots. Sutton (1991) examined some of the outcomes of this debate.
A further aspect of Baumol’s contribution is his concept of efficient component pricing, itself a
development of the earlier work on contestability and pricing rules. Here, Baumol and Sidak (1995)
sought to show that pricing of services sold by a vertically integrated supplier to a competitor in one of
the vertically-related markets should be at a price which did not reduce the profits of the vertically
integrated firm. This is also controversial, but could well have relevance to the vertically integrated
activities of major players in the food processing and distribution chain. An example might be where
the integrated retailer wholesales to independent retailers. The wholesale pricing rule should, under
this rule, be such that any diversion of market share would not reduce the integrated firm’s profits
below what could be expected without the wholesaling activity. This could become of interest to
public policy if the wholesaler happened to be a natural monopolist, or possess some degree of market
power, in part of the total independent retailer market. This efficient components rule seems on the
surface to be a licence to holders of market power to continue to enjoy that power: the efficient
components price is the price which would make the seller indifferent between selling to an outsider
firm or engaging in the activity itself. Therefore, the powerful, supernormal profit earning seller will
continue to earn those profits on capacity retailed to an alternative supplier of the downstream output.
Amazingly, this rule was approved by the New Zealand courts in relation to NZ Telecom selling
capacity to Clear, the entrant in long distance telephony (King and Maddock 1996). It may not be
entirely irrelevant that Baumol and his colleagues worked at Bell Labs, a part of AT & T which has a
dominant position in long distance and international telephony in and from the USA (Shepherd 1995).
Again, the relevance to our context is the integrated wholesaling and retailing firm, with activity in a
number of different retailing formats (supermarket, discount stores, specialist shops), and which
contests intermediate markets, wholesaling to independent retailers. The analysis of the profit
maximising choices which such a firm might make would suggest that the efficient components rule,
combined with Ramsey discrimination, should be used as an hypothesis to be tested. If this were to be
substantiated, the next question is whether this is the exercise of market power. If so, is it
monopolistic behaviour of interest to the ACCC, or is it something for which any other remedy could
be justified?
The testing of these ideas with econometrics has done little more than not contradict the impressions
of less than casual observers (Scherer and Ross, 1990, Ch 4), and the generalisations are few and not
very strong (Sutton, 1991, Ch 14). Unfortunately, the New Industrial Economics has left
generalisation even further behind (Brander, 1992). Game theory has been fruitful in ex post
rationalisation, but relatively fruitless in ex ante prediction, and incapable of any but the most obvious
generalisations, ones made decades earlier by the obsolete "old" industrial economics. But ex post
rationalisation may be a useful contribution in trying to distinguish between credible and non-credible
entry threats in order to explain differences between the exercise of market power and fierce rivalry in
the context of case studies. Sutton (1991) concluded that the fundamental outcome of this work was
the significance of sunk costs as the token of credibility. Commitment to investment is the only way
in which entry or maintenance of a market position can be taken seriously by incumbent or potential
competitors. It is to be expected that financially powerful firms will invest in directions which will
enhance their position. Recent contributions to industrial organisation theory make much of the
strategic aspects of firm decisions. Strategic decisions can be identified by the effect they will have on
raising further barriers to entry. The main evidence is that firms invest in directions that increase the
level of sunk costs required for profitable operation. An example is investment in firm-specific
network capital goods, such as Just-In-Time delivery systems, that have to be implemented backwards
in the supply chain. Whether the "contestability" label was more than a convenient method of
summarising these results or a real theoretical advance remains controversial (Shepherd 1995).
However, it has allowed the empirical economics of markets to devise markers for departure from
competitive performance (see Chapter 3 and Appendix B).
11
2.6 New Institutional Economics
Neoclassical theory sees the firm as an entity with a set of single minded goals, without any agency
problems, information difficulties or uncertainties about markets or production possibilities. But all of
these information problems have important implications for behaviour and for the exercise of market
power. The New Institutional Economics sees these problems as the rationale for the existence of the
firm and its particular form within its markets. Putterman & Krosner (1996) re-produce the most
important papers in this tradition. Ronald Coase (Coase 1937) is the seminal author in transactions
costs economics. He pointed out that the market has costs, and that the firm may substitute for the
market where transactions costs of the market can be avoided. This is, at bottom, a rationalisation for
vertical integration. But, more than that, it suggests that organisation is a response to economic, or
competitive, pressures. The nature of the institutions, of firms, markets and of all the various forms
with elements of both, is an outcome of economising choice, and not some exogenous chance.
Agency problems exist where two parties to a transaction have different goals, or different
information, and are intent on achieving their own goals regardless of other parties. These problems
can be found in transactions between firms, as when a producer and a buyer have different perceptions
of the situation. The producer may know something about the quality of the product which is
unknown to the buyer; the buyer may know something about the market in which the transaction is
taking place which the producer does not know. This is known as information asymmetry. Either
may hide information from the other to achieve their goals. Similar problems may be seen within an
organisation when, for example, the goal of a supermarket produce buyer is short term profit
contribution, while the strategic management goals of the supermarket focus on long term supplier
loyalty. This can be a result of poorly designed incentive schemes or poor organisational information
channels. These are both symptoms of poor organisational design in general.
Agency and transactions costs problems can help explain particular organisational forms and
structures. The nature of firms, alliances, franchises, contractual arrangements, markets and of all the
various forms they take, is an outcome of economising choice. The firm is organised to enforce
performance by opportunistic input suppliers in a context of information asymmetry, where no single
entity can observe all aspects of performance. According to the literature in this area, the institution of
the firm, and its financing, is a result of a choice of institutional form to minimise the costs of
monitoring performance. The relevance to the food distribution chain is clear: one observes in that
chain many types of relationships from vertical integration to arm’s length markets. Independent retail
stores band together in marketing groups, often co-ordinated by a wholesaler. Retail chain stores
integrate some functions but outsource others; they also have contractual relationships of various
kinds, formal and informal, with their suppliers. “Own brand” producers are tied to retail chains or
wholesalers by contracts which vary in the exclusiveness they demand of the producers. Processors
and manufacturers also integrate backwards and horizontally to form the large, sometimes
multinational, grocery and commodity companies. They also have supply contracts with farmers and
other suppliers which are not simple arm’s length market relationships but have a long life, and
expectations on both sides of continuity, no matter whether they are more or less informal.
Problems in these relationships are suggested by some of the submissions to the recent Joint Select
Committee on the Retail Sector inquiry (Chapter 4). Contractual relationships between growers and
processors or retailers vary in formality from mere handshakes to detailed legal agreements. In both
cases there have been assertions of agreements being interpreted or varied in favour of the party with
the greater financial power, be it a chicken processor or a retail chain. These agreements and their
evolution may be interpreted as a market response to a need for two things. The first is efficiency
improvement (Just-In-Time and Total Quality Management are two slogans used in this regard in the
business world). The second is for the "market makers" to construct a system which serves their needs
in terms of allocating risk, preserving proprietary knowledge advantages and avoiding adverse
selection of subsidiary network members. For example, an agreement between a retail chain and a
broccoli grower may be silent on the question of whether the grower's price can be decreased because
of marketing effort (e.g., discounting) on the part of the supermarket. Evidence was submitted to the
12
retail sector inquiry that the retailer discounted the grower’s price when the retail price was discounted
without consultation with the grower (NFF 1999 and QFVG 1999). Does this sharing of the
marketing cost lead to improved efficiency or is it a mere shifting of profits? Analysis of the agency
relationships involved, in conjunction with production technologies, should allow a statement about
whether the practice is merely the exercise of market power or the adjustment to more economically or
technically efficient production.
2.7 Evolutionary Economics
Neoclassical economics, including much of the New Empirical Industrial Organisation Economics,
leaves important aspects of economic behaviour exogenous. Investment decisions, technology, and
the choice of activity for a firm are all within the "black box". Institutional economics looks at the
organisational forms which emerge where there is a competitive capital market and an assignment of
property rights, but ignores questions of competitive advantage. Evolutionary economics examines
the competitive process rather than the equilibrium conditions of the mature market, to explain
technological trajectories, market structures, variety of firms, products, processes and communities of
industries, in much the same spirit as ecologists and evolutionary biologists. The business
management literature has its own version of this literature, the resource based theory of business
strategy (Ansoff, 1965). More recently, these literatures have come together at least in management
journals. In economic theory, the work of Penrose (1959), Richardson (1960) and Nelson & Winter
(1982) has been seminal. Penrose analysed the growth of the firm as a process of exploiting and
generating specific resources, especially human resources. Richardson examined the problem of
market knowledge in co-ordinating investment by firms related by market interactions, and found that
"market" and "organisation" were very inadequate descriptors of what is necessary for that coordination. This is especially relevant to the food chain of tightly linked firms, whose interdependence
is more extreme than is the case for some other sectors. Nelson & Winter formalised an evolutionary
model where the firm was characterised as a set of production and investment routines and a set of
routines to modify the production routines. Their model, a development of Marshallian and
Schumpeterian theories, has become the basis of an extensive literature which encompasses much
more than the neo-Marshallian/Schumpeterian model which began it.
The marketing and management literature has a strong interest in this area, as shown by the
importance of papers in this tradition for journals such as Strategic Management Journal, Journal of
Marketing, Academy of Management Review, Journal of Management Enquiry and Industrial and
Corporate Change. Edith Penrose herself wrote her seminal work in part as a case study of the
Hercules Powder Company, an important work in business history. The recently coined term,
Resource-Advantage Theory of Competition draws heavily upon her work, explicitly as well as
implicitly. It is precisely her insistence on the unique nature of the resources available to the firm that
can be characterised as "Austrian", a characterisation eagerly grasped by the Resource-Advantage
writers. But, unlike Hayek (1945) and Kirzner (1973) who see entrepreneurial activity as leading
markets to equilibrium, these writers follow Penrose in seeing market forces as doing more than
merely equilibrating demands with supplies. The seeking of opportunities which drives firms is
anything but equilibrating, in her view. It is creative and sometimes destructive, in the way
Schumpeter, another Austrian, saw competition.
Like the paradigms based on classical and neoclassical economic theories, evolutionary theories
examine the production of outputs, rather than focussing on optimising exchange relationships.
However, exchange cannot be ignored, given the role of private knowledge in the competitive process.
Richardson’s insights into the entrepreneur’s need to create sufficient certainty to allow investment
gives the necessary rationale for organisations to form and reform. Vertical and horizontal integration,
and the multitude of relational forms which are found in complex production systems, are explicable
responses to the uncertainties which the impersonal market would pose for would be producers.
Knowledge is central to these theories, the knowledge of how to produce outputs. This is specialised
to the firm in question. Each firm has its own fine variant on doing what it does, even within a
seemingly homogeneous industry. Knowledge advantages create the profit the firm can make. If all
13
knowledge were public, no profits could be sustained, so no investment would be worth carrying out
(Richardson 1960). The competitive process would immediately grind all activity to a stagnant
equilibrium in which no further investment or innovation would be undertaken. This dreadful picture
is so far from the capitalist economy that it can be dismissed. Private knowledge drives
entrepreneurial activity, but private knowledge also leads to potential for the exploitation of market
power. Here one can also see that market power and the competitive process may be considered to be
interdependent, or symbiotic, both being outcomes of the drive for profit that is the capitalist
economy. Nightingale (1996) has reviewed this literature.
The power of the resource based evolutionary theories is in their ability to predict and explain changes
of market structure and directions of technological change. The path dependence of market
development is central; history is not merely the learning of lessons from the past, but a means of
pointing toward the future. Technological directions are seen in lock-in effects of past investment and
in strategic interests of firms whose future goals are to entrench their knowledge advantages. The
dominant large firm invests in increasing economies of scale and scope; the innovative entrant invests
in techniques which subvert the dominant firm’s position, and transform the selection process of
market success or failure.
The observation that profit rate and other performance variables differ more between firms than
between industries gives some indication of the importance of private knowledge in the capitalist
system (Scherer and Ross 1990, p. 650). Orthodox economic theory does not cope well with this
observation. Evolutionary or resource advantage theories expect this to be the case, and explain why
it is so.
The relevance of resource advantage and change trajectories to the present context is in the growth and
technological development of the food marketing chain. The investment in innovation, and its
diffusion, which drives the firm in its quest for long run profit and growth, is possible only where the
firm is sufficiently profitable and sufficiently pressed by forces of competitive rivalry. These two
conditions, the financial resources and the incentive, are the product of market processes, the selection
mechanism transferring market share from the less efficient to the more efficient, and the innovation
mechanism generating new varieties of technologies and products. Market power is both an incentive
to growth and innovation as well as a long term end to the selection process if innovation fails
systemically. Concentration increases if efficient firms continue to accrete market share without
check from innovative moves by previously less efficient firms or from entrants.
The directions of change under these theories are conservative and predictable for most periods of
time, as firms search for new efficiencies within their existing technologies. This means it is possible
to make reasonably confident predictions about general directions for growth, within a context of
technological forecasting. What is not possible is to predict change that is radically creative. One can
be sure there will be such change sometime in the future, and that it is likely to be associated with
successful entry or the equivalent of entry by reconstruction of an existing firm. The last time the food
chain was subject to such a change within Australia was the entry of the variety chain stores,
Woolworths and Coles, into grocery retailing in the early 1960s. Before that, the cut price self-service
methods of Flemings, Franklins and Tom the Cheap Grocer, to name the three most "notorious"
players, transformed the way groceries were sold in Australia during the 1950s. Even these changes
were, at the time, generally predicted from knowledge of retailing elsewhere in the world.
In the coming decade the predictable trajectories of change include global horizontal integration of
retailing systems and tighter vertical linkages in the supply chain, the change causing most angst for
suppliers here in Australia. Not much more than educated commonsense is required to guess that
further entry to food retailing is likely to occur by firms with innovative techniques. More than
commonsense is needed to examine the implications of alternative scenarios following entry. This is
where evolutionary modelling will be used by stakeholders in the industry in their efforts to second
guess their own futures.
14
By way of digression, it should be noted that Michael E. Porter’s concept of competitive advantage
(Porter, 1980, 1985, 1990), while couched in the language of SCP, is a manager’s guide to creating
and exploiting resource advantage. Any situation may be fitted into Porter’s framework: it is a
taxonomy that allows consideration of the complete set of opportunities and threats facing a firm, from
the five directions of his diagram. The firm is encouraged, by his framework, to follow a competitive
strategy to take advantage of the situation it is in. When will the firm choose a strategy which might
contravene the Trade Practices Act? When will a strategy be an attempt to exploit market power? Here
there is another use of the taxonomy; namely, to explore the various possibilities his paradigm shows
us. If one wishes to communicate effectively with business managers it is probably the case that this
framework should be employed, at least as a "user friendly interface".
2.8 Policy Implications
All the above suggests that market power is a complicated and difficult concept to tie down. How is
social policy to make fine judgements about markets when it appears that economists have such vague
views about identifying market power in practice?
Each of the theoretical perspectives reviewed above suggests its own particular indicators and
remedies. Neoclassical theory suggests examining price-cost margins, the existence of conjectural
variations that suggest firms react to each others’ actions (see Appendix D) and inter-industry rates of
profit. This theory is a theory of market behaviour, rather than of firm behaviour. It is appropriate
where it is sensible to expect equilibrium observations and statistically tractable data sets. Where
either of these is not to be expected, alternatives should be taken. Either large numbers, or monopoly
(or tightly disciplined cartel) is the requirement.
The SCP approach takes a broader view of the phenomena to be examined. In particular, this
approach sees as perhaps of much greater significance than mere price "gouging", the use of market
power to retard innovation and subsequent productivity change. The market system should result in
resources going to those who can best use them, but if powerful players can retain resources in spite of
more efficient firms, then society is damaged by the loss of real income over time: the growth path is
shifted downwards. Such a downward shift of the stream of real income constitutes the real loss to
society. It is arguable that the poor performance of the Soviet economy over the post WWII period
was mainly due to failure of resources to respond to possibilities for innovation and cost efficiency
over the long term, rather than simply to poor resource allocation in a static sense. The costs of
technological lock-in are an example of losses due to misallocation of resources at some point in time
leading to a lower productivity level at every subsequent point in time. Technologies such as the
internal combustion engine and the QWERTY keyboard may be dominant due to lock-in effects
(David 1985). Those and other examples may have resulted from exercise of market power. The
question is whether the costs of change outweigh the benefits under a social benefit cost analysis. But
is the standard analysis, mentioned above, appropriate where the future path of income may be
improved to an extent which is radically uncertain? What if the costs of change are all in the
immediate future, but the benefits increase exponentially over time? At some high rate of discount the
change would not be beneficial, despite uncertain expectations of rapidly increasing benefits in the
further distant future?
This highlights the fact that much market power is transitory. The longer the perspective, the less
significant is that power. How long is too long? This depends on the social rate of discount, and the
extent of the losses per unit of time. Bringing the losses back to a present value may give the policy
maker an idea of the cost that would be worth incurring to ameliorate the damage. This is the
approach taken by the old "workable competition" literature, associated with the SCP approach
(Sosnick 1958): take action if the cost of acting is less than the expected benefits. The onus is then on
the competition authority to establish what the likely costs and benefits might be. This is typical of the
policing functions of a regulatory authority of any kind, and raises the question of search procedures
highlighted by Simon’s bounded rationality conundrum: how do you set a STOP criterion for the
15
search for information where the cost of search is known, but the value of the findings is unknown
before they are discovered (Simon 1966).
Evolutionary theory and the struggle for competitive advantage is also seen in the rate and direction of
technological change. Evidence that this rate is being repressed by market structure is only one
element, highlighted by SCP’s concern with overly high concentration with high entry barriers. The
direction in which change is headed (for example, toward the creation of higher entry barriers) is one
indicator of the exercise of power. If firms enjoying current market success succeed in hard wiring that
success by strategic choices to prevent followers from enjoying the same success, the market process
is subverted. Evolutionary theories suggest looking at the firms undertaking the R&D or the
investment in innovation to understand the implications of this for future competitive processes. The
historical prospect of entry and exit in relation to profitability of leading firms is another indicator of
the health of the competitive process. Persistent high profits without entry and without the churn of
market share amongst the leading firms would be an indicator of market power, according to an
evolutionary perspective. The recent history of grocery retailing is one of some churn, and of some
fluctuation of profits and market share, with Coles, Woolworths and Franklins each experiencing good
times as well as bad. Current plans for entry by Aldi would suggest that the current era of dominance
by the supermarket may come to an end. Entry barriers would seem to be high enough to deter entry
by start-up firms lacking "deep pockets" to finance a long establishment period, but not high enough to
deter the well financed entrant (Aldi is said to expect to lose money for the best part of a decade; see
Appendix A).
A significant question, motivating concerns that led to the Joint Select Committee on the Retailing
Sector, is the exercise of power via the information asymmetry between parties, which creates an
agency problem for the market, as discussed above. Evaluating the significance of this asymmetry of
power is not easy. To what extent should society be concerned that market rents are being redistributed from some entities to others? If the viability of the rent-losers is affected it might seem a
cause for intervention. But what rent-gainer would kill the goose laying the eggs, even if not quite
golden? It would seem to be a serious error on the part of the gainer. Equity issues may be subject to
Part IVA of the Trade Practices Act, unconscionable conduct. To bring such action is the
responsibility of the injured party, rather than of the ACCC. Sustained and general conduct of this
kind may be dealt with under Part IVB, industry codes, and suggestions have been made along these
lines by the Select Committee.
Similar issues arise when the question of interest is the distribution of the costs of market power. If
the market power is exercised against small firms (the large scale technology bias case) but actually
benefits consumers, despite the higher level of concentration spawned by that market power, then it is
very hard to argue to remove that power. Only if there is strong evidence that there exist small scale
technologies that are likely to be more cost efficient in the long run will social policy be inclined to
move against the sources of market power. The prima facie evidence is that consumers have benefited
from increasing scale, and that competitive forces have constrained incumbents at the retailing end of
the chain, and ultimately not prevented entry. Action to limit the ability of large scale firms would
therefore seem not be to justified by this evidence.
To sum up, economic theory does not suggest any “blue print” for assessing the presence and extent of
market power. Rather, it suggests some things to look for which might be conducive to firms being
able to earn above-normal profits for a sustained period of time. These include high levels of sunk
costs, particularly those associated with marketing and, to a lesser extent, those associated with scale
enhancing technological change, and asymmetric information. The presence and extent of market
power has to be assessed on a case-by-case basis. Some firm characteristics such as innovative flair
(cost-reducing technologies and new product development) are a means by which firms can make
above-normal profits over a relatively short period of time. But these characteristics are not only
desirable in the eyes of consumers but are essential to a capitalist economy.
16
3: Insights from Empirical Studies
In this chapter selected empirical studies are reviewed. The aim is two-fold: to take stock of what the
empirical studies have revealed about the existence of market power and to demonstrate the range of
methods that have been used. There is some discussion of studies done on different continents.
3.1 Traditional Structure-Conduct-Performance Studies
In the traditional structure-conduct-performance paradigm, the emphasis was primarily on a one-way
causal flow where the structural characteristics of an industry determined the conduct or behaviour of
firms in that industry, which then determined measures of performance such as profit rates (Bain
1951). In such a paradigm, conduct was a consequence of structure and therefore did not have much
independent interest. Performance could be predicted by structural characteristics alone. Thus the
focus was principally on cross-section studies of many industries, and whether profits were higher in
more concentrated industries. Coincidentally, quantitative measures of structural characteristics and
performance variables were fairly readily available while numerical measures of firm conduct were
much more difficult to come by.
There have been a huge number of empirical studies which have estimated these sorts of models.
Performance measures were typically accounting profit rates or price-cost margins. Structural
characteristics of industries were typically 4-firm or 8-firm concentration ratios, with controls for
other elements of industry structure (and conduct). The earlier studies tended to be quite parsimonious,
while later studies added more and more control variables. Most studies were done in the heyday of
the traditional industrial organisation paradigm in the 1960s and 1970s and early 1980s, and most
were done in the US. As early as 1974, Weiss (1974, 193) conjectured that it “..must be one of the
most thoroughly tested hypotheses in economics by now.” Reviews of this general sector-wide
empirical work have been fairly few but seminal (Weiss 1971, 1974 and 1990, Vernon 1972, Scherer
1980). Reviews in a agricultural or food sector context were undertaken by Helmberger et al. (1981)
and Griffith and Gill (1984).
Weiss (1974) tabulated 46 mainly manufacturing sector-wide studies and cited another eight (his
Table 11). He concluded that the large majority of studies showed a significant positive effect of
concentration on profits or margins, and that this relationship was quite robust across different time
frames, countries, measures of structure and performance, other variables controlled for, units of
observation, data sets and data sources. In those studies which used the preferred price-cost margin as
a performance measure, rather than rate of profit, the estimated relationships were almost all positive
and significant. North American sector-wide studies along the same lines but completed after the
Weiss survey, say in the period 1975-1985, generally confirm the earlier findings.
Several studies during this period examined performance in food retailing. A general conclusion from
these studies would be that markets which had small numbers of larger retail firms were more
profitable than markets which had larger numbers of smaller retail firms.
Several Australian studies in the 1960s and 1970s followed the US literature and examined the
standard concentration-profits relationship in the context of the domestic manufacturing sector. Again,
most were at the sector-wide level. For example, Norman (1971) concluded that Australia’s more
highly concentrated industries had performed better than less concentrated industries in terms of
growth of output, technical progress and changes in average price levels. Conversely, Sheridan (1974)
suggested that there was no relationship on a single year basis between firm size and profitability in
Australian industries.
Round (1975) found an insignificant relationship between average industry profit and concentration in
33 Australian manufacturing industries. However he found at higher concentration levels that there
17
was a significant relationship with profits, that large firms improved their profitability compared to all
firms in that industry, and that the spread of profit rates between large firms significantly widened. His
conclusions tentatively supported the hypothesis that rates of return were higher in large firms because
of greater efficiency, rather than collusive conduct or abuse of market power.
At a more disaggregated level, little empirical research under the traditional paradigm was undertaken
in Australia except in relation to the meat industry. All of these were time series rather than crosssectional studies, and based on ad hoc theoretical specifications. For example, Griffith et al (1984)
investigated whether the changing structure of pigmeat marketing in the early 1980s had any impact
on pigmeat price spreads. Concentration variables were not found to have any consistent or significant
separate impact on price spreads in pigmeat marketing. In a similar vein, Corbett (1998) recently
examined whether the rising proportion of beef sold by supermarkets relative to butcher shops in
NSW, as a measure of increasing concentration in meat retailing, was able to explain any of the
increase in the beef farm-retail price spread. The concentration variable had a positive sign in most
model specifications but was generally insignificant.
There were a major deficiencies with the types of studies quoted above. The evidence that higher
concentration caused higher profits said nothing about the source of the increased profits, ie nothing
about the behaviour of the firms making up the industry and whether they were exerting some sort of
market power or whether they were just more efficient. This criticism led to a renewed interest in firm
conduct and the development of the New Industrial Organisation as promoted by Porter and others.
3.2 New Empirical Industrial Organisation Studies
The focus of the New Industrial Organisation literature is on the conduct of firms within a particular
industry, where the industry is allowed to depart from the competitive model. The New Empirical
Industrial Organisation is an attempt to continue to use systematic statistical evidence while returning
to the study of single (or closely related) industries and focussing on firm behaviour directly.
Innovative econometric models are used to estimate parameters of industry marginal cost functions
and measures of industry conduct using conjectural elasticities, based on various theories of
imperfectly competitive markets. The NEIO literature up to the late 1980s has been surveyed by
Geroski (1988) and Bresnahan (1989).
Thus behavioural equations by which firms set price and quantity are estimated and the size and
significance of particular parameters infers coherence with a particular theoretical explanation of
interaction between firms. Bresnahan (1989, pp.1014-1018) provides a stylised industry model in
algebraic form; O’Donnell (1999) provides a Thus behavioural equations by which firms set price and
quantity are estimated and the size and significance of particular parameters infers coherence with a
particular theoretical explanation of interaction between firms. modern perspective of the algebra.
In one form, popularised by Applebaum (1979, 1982), conduct is described by the conjectural
variations of firms in an industry, that is their expectations about the reactions of other firms to an
increase in quantity. Profit maximising behaviour is assumed to describe the underlying model of firm
conduct operating in that industry and first order conditions are derived which include measures of
these conjectural variations. A system is jointly estimated which includes the industry demand
function, the firm equilibrium condition and/or the factor demand functions. Thus the measures of the
conjectural variations parameters are obtained from actual data. Early examples of this approach
include Roberts (1984) and Schroeter (1988), while recent examples include Muth and Wohlgenant
(1999) and Park and Weliwite (1999).
O’Donnell (1999) uses this sort of approach in the context of a marketing margin model. He takes a
general industry model, imposes restrictions consistent with some specific assumptions about fixed
proportions, linear demand, normalised quadratic costs, and constant returns to scale, and generates an
estimating equation for the farm-retail margin. Coefficients on the quantity variables in the models are
functions of the conjectural variations parameters in input and output markets. A test of whether these
18
coefficients are significantly different from zero is equivalent to a test of whether the conjectural
variations are significantly different from zero. If they are, market power is implied; if they are not
competitive behaviour is implied.
In another form of these models, the specification is based on a particular theoretical framework, the
critical parameter to be estimated is θ, the industry conduct measure, or more correctly the average of
the conduct parameters of all firms in the industry. θ = 0 implies perfect competition, θ = 1 implies
monopoly, and different values of θ in between will be predicted by different oligopoly theories. An
example of this approach is Holloway (1989). Restrictions on parameter values are derived which
imply competitive or non-competitive behaviour. Statistical tests are done on the restrictions. If any of
the F-tests for the restrictions, that the demand shift and supply shift variables be equal but of opposite
sign, are significant, this implies at least one piece of evidence that the industry is not competitive. On
the other hand, if none of the F-tests are significant, the empirical data are consistent with competitive
behaviour. Zhao et al (1997) apply this sort of model to Australian data.
Baker and Bresnahan (1992) provide a non-technical description, and Bresnahan (1989) provides
extensive detail, of the ways that various studies have gone about inferring the causes of any market
power which has been identified through estimation of industry conduct or conjectural variations
parameters. Bresnahan summarises “In any particular industry, the available information and
institutional detail allows different kinds of analysis and different defences of different analyses. One
can therefore expect some continuing variation in desired method.” (p.1045). He further points out the
additional difficulties posed by changes in industry structure over time, and by product differentiation.
3.2.1 Examples of US and Canadian Studies
Bresnahan (1989) compared the estimated price-cost margins from various and Canadian studies
covered in his review. The estimated price cost margins ranged from about 5% to about 90%. These
studies covered manufacturing, retail and service industries. He drew three conclusions from that
review:
•
there is a great deal of market power, in the sense of price-cost margins, in some concentrated
industries. He is cautious about being more definite in this conclusion because of two reasons.
First, there is the problem of self-selection of industries to study - authors testing new methods
wish to show they work by applying them to industries where it is highly likely that anticompetitive conduct exists. Second, because of the nature of the industry data required for these
types of models - industries in which information about what competitors are doing is quite good
might well provide an inference of collusive arrangements.
•
one significant cause of high price-cost margins is anti-competitive conduct. Some of the studies
found conduct well towards the collusive end of the spectrum, but there were substantial
differences between firms in some industries. Eg between dominant and fringe firms.
•
only a very little has been learned from the new (NEIO) methods about the relationship between
market power and industrial structure. Bresnahan is pointing out here that the studies have mostly
focussed on the concentrated end of the industrial spectrum, and that even though market power
can now be more easily and consistently measured, still not very much is known about the causes
of market power. He points to the need for more work on the determinants and effects of entry
into concentrated industries, and on long run strategic behaviour by firms including predatory
pricing.
Food Manufacturing
There have been many US studies using these methods with a direct focus on agriculture/food
markets. With respect to food processing, the meat packing sector has been the focus of a substantial
amount of research activity in the last decade or so. Many found market power in the purchase of
19
finished cattle and/or in the sale of packed beef. The Koontz et al (1993) study was different to the
others in that is was based on a non-cooperative game theory model. Evidence of both cooperative and
non-cooperative conduct was found in each of the four markets studied, thus indicating monopsony
buying power, but this conduct did vary across the markets and over time. Thus firm conduct in these
markets was thought to vary as industry supply and demand conditions changed, and as the market
structure adapted to higher or lower profits during particular periods, and firms entered or exited the
industry.
However recent research casts doubt on these findings. Muth and Wohlgenant (1999a) found no
evidence of market power in this industry in either input or output markets. Muth and Wohlgenant
(1999b) continued this line of inquiry with a more flexible modelling approach similar to Bresnahan
(1989). This was a two-equation model which includes the demand function facing the industry and
the industry equilibrium condition where marginal cost equals perceived marginal revenue. This latter
equation includes a parameter which is an index of market power, and this is allowed to vary in
response to changes in industry conditions (concentration ratio, input costs). Thus the degree of
market power is estimated from actual data. When the market power terms are held fixed, imperfect
competition is found in at least one market. However when the market power terms are allowed to
vary with changing industry conditions, no evidence of market power is found.
In a series of influential papers, Paul (1999a,b) stressed the need for a rigorous treatment of the cost
structure of the industry when attempting to measure market power effects. In (1999a), she used a
cost-oriented model of cost economies and market power and meat packing industry data at two levels
of aggregation - aggregate industry data, and plant level data for 42 separate firms. Three cases were
studied and consistent results were found: if monopoly or monopsony was acknowledged, market
power was found to be significant but not very large; if both monopoly and monopsony were
acknowledged, net market power was found to be significant but it was difficult to separately identify
input and output market power; while if the plant data were used significant market power was found
for both beef output and cattle inputs. However, significant scale and other cost economies were also
found, and these tended to balance out the market power effects. Paul concludes (p.629) "Increasing
concentration in the US meat packing industry seems justifiably to have emerged from cost
economies, which appear in turn to be primarily transmitted to suppliers and demanders of cattle and
meat products rather than generating excessive profits for the plants or firms."
Food Retailing
With specific regard to food retailing, Holloway (1991) found no major departures from competition
in the whole farm-retail marketing chain for eight major US food groups, Ailawadi et al (1995) found
no concrete evidence confirming an increase in market power exercised or accumulated by grocery
retailers, and Messinger and Narasimhan (1995) found that neither accounting nor stock market data
clearly indicated a shift in channel profitability from manufacturers to retailers.
In the most recent and most rigorous treatment, Park and Weliwite (1999) use aggregate retail industry
data obtained from official and trade sources to examine whether there has been any evidence of
market power in US food retailing. They use a two-equation model based on Bresnahan (1989) which
includes the demand function facing the industry and the industry equilibrium condition where
marginal cost equals perceived marginal revenue. This latter equation includes a parameter which is an
index of market power, and this is allowed to vary in response to changes in industry conditions and
technical innovations. Thus the degree of market power is estimated from actual data.
They find price taking behaviour in US food retailing prior to 1983, but some evidence of market
power since then following an increase in merger activity. The index value was estimated to be 0.018,
suggesting that the average level of competitive industry output since 1983 would be at most 1.8%
higher than the average actual output. A problem with these results though is that the measure of
increased merger activity is just a 0/1 dummy variable, so there is no association possible between
incremental changes in structure and increases in market power. Further, neither the adoption of
20
scanning technology nor an increase in firm size (sales/checkout) were significant contributors to the
index.
Wohlgenant (1999) provides an updated review of the various models researchers have used to
examine food marketing margin behaviour, including the behaviour of non-competitive components of
the marketing chain.
3.2.2 Examples of European Studies
There is little empirical evidence available on the issue of market power in European food marketing
channels. In 1989, the UK government enacted measures that led brewers to divest themselves of
some 14,000 hotels. Since then, beer prices have risen. Slade (1995) develops an econometric model
of the transition period based on the relationship between retail price and retail organisation form that
emphasises the strategic aspects of the relationship. The estimated relationship implied that the
divestiture resulted in cost diseconomies.
Finally, the UK Office of Fair Trading has recently completed an inquiry into food retailing (OFT
1999). On the basis of their findings, the Office has referred a formal competition inquiry to the
Competition Commission. The four reasons given for the referral were that the level of profitability of
the four largest chains were high, that there were significant barriers to new competitors, that grocery
prices were often set to match competitors rather than undercut them, and that suppliers including
agricultural producers were adversely effected.
3.2.3 Examples of Australian Studies
In the meat industry, Hyde and Perloff (1997) found that the domestic retail meat market was
competitive for beef, lamb and pork and that market power had not increased over time. Zhao et al
(1997) modified the model developed by Holloway to account for trade and applied it to the
Australian beef market. When the domestic and export markets were separated, there was no evidence
found of non-competitive behaviour in the domestic beef market.
A particular area of concern in Australia in recent years has been retail fluid milk markets, and the
extent to which retail prices might change because of concentration in food retailing following
deregulation in the various states. For the product of primary interest, carton milk, in the retail to
farmgate equation, O’Donnell (1999) found that the estimate for the term containing the conjectural
elasticity on output was significantly different from zero, implying that in deregulated markets carton
milk retailers do possess market power in the sale of market milk to consumers. Similarly, evidence
was found of carton milk processors possessing market power in the purchase of market milk from
farmers. These results were confirmed in the processor to farmgate equation, where carton milk
processors were found to have market power in both output and input markets. While significant
evidence of market power was found, O’Donnell (1999) was unable to estimate the actual conjectural
elasticity parameters because of data constraints. Thus the general result implies market power in the
market milk processing and retailing sector, but the extent of this market power (the value of the θ
parameter) is unknown, as is the cause.
3.3 Time Series Methods
Another recent theme of research has been to examine the time series properties of the major price
series in a particular food chain to infer industry behaviour from market outcomes. For example, Reed
and Clark (1998) used cointegration methods on data from seven US food industries. These data
contained both deterministic and stochastic trends, and the author's focus was on the effects of
different ways of handling these trends. They found that ignoring the stochastic trends leads to a
rejection of the competitive model hypothesis for six of the seven industries (like Azzam and
Pagoulatus 1990); that removing these trends by differencing leads to a rejection of the competitive
model hypothesis for none of the seven industries (like Wohlgenant 1989); and that properly
21
accounting for them leads to a rejection of the competitive model hypothesis for only two of the seven
industries. Thus analysts are more likely to reject competitive market behaviour if they do not
correctly account for the characteristics of the data series they are using. These authors also make the
point (p.1142) that "... deciding whether markets are competitive rests not on whether an industry
establishes a gap between price and marginal cost but on whether the gap is maintained over time and
as capital moves in and out of the sector."
Goodwin and Holt (1999) also used cointegration methods to examine the long run equilibrium
relationships between US farm, wholesale and retail beef prices, with a particular emphasis on the
causal direction of price changes and on whether responses were symmetric to price rises and price
falls. They found unidirectional price transmission, from farm to wholesale to retail, which does not
imply the existence of market power at higher levels of the chain. They also found that the
responsiveness to price shocks had increased in recent years and they suggested that this result may
imply that markets have become relatively more efficient in transmitting information through vertical
marketing channels.
Palaskas (1995) used cointegration methods to examine the dynamic transmission of five agricultural
producer prices through the food marketing system in seven countries of the European Union.
Statistical tests rejected the hypothesis of short run perfect price transmission across all pairs of prices
in all countries. Further, in the long run, perfect price transmission was not rejected in only two of five
price pairings across six of the countries. The UK was the exception. So for all products in the UK
and for three of the products in the other six countries, non-competitive behaviour was concluded.
Dawson and Tiffen (1997) also use cointegration models to look at beef, lamb and pork farm and retail
prices in the UK and Wales. They could not find a cointegrating relationship for either beef or pork
and while they did find one for lamb, the relationship implied causality from retail to producer prices.
This suggests that for the lamb industry, prices are set at the retail level and thus there is some
evidence of market power in retailing, while for the beef and pork industries there is some evidence
against a competitive market. A problem with both analyses is the confounding impacts of agricultural
policy measures which are in place to influence and supplement farm prices. These impacts are not
properly handled in either of these studies.
In Australia, Chang and Griffith (1998) found that the farm, wholesale and retail prices for beef were
cointegrated and thus moved together over time, all responding to exogenous shifts in demand and
supply curves. However the wholesale price was found to be weakly exogenous, suggesting some
disequilibrium behaviour.
3.4 Conclusions
What does the empirical evidence tell us about the relationship between concentration levels,
consumer prices and prices received by farmers?
The first point to make is that very few empirical studies of the type reviewed here have been done in
the Australian food marketing chain. Three manufacturing sector studies done in the 1970s were
inconclusive, several studies of the meat industry have suggested competitive conduct while two
studies of the dairy industry have indicated non-competitive conduct (although there are policy
interactions in this industry). For a broader perspective one has to look to studies done in North
America and Europe for guidance.
Recall that in the US the manufacturing sector tends to be the chain leader, although that may be
changing, so most policy attention has been on food processing and manufacturing industries. The
conclusion drawn from the older US structure-conduct-performance literature was that the large
majority of studies showed a significant positive effect of concentration on profits or margins, and that
this relationship was quite robust across different time frames, countries, measures of structure and
performance, other variables controlled for, units of observation, data sets and data sources. Thus,
22
industries which had small numbers of large firms were more profitable than industries which had
larger numbers of small firms. Similar conclusions were made for food retailers.
Conversely, the conclusion drawn from the US New Empirical Industrial Organisation literature is that
no such generalisation is possible. Certainly there is evidence of non-competitive conduct in some
manufacturing industries, but the evidence also varies by which industry is studied (since there is an
element of self-selection), by definition of market (national or regional) and over time (as market
conditions change). There are several studies which produce opposite conclusions for the same
industry over the same data period. At the retail level, the majority of studies find competitive
behaviour, although the most recent study by Park and Weliwite does conclude that there is a small
but significant market power premium in US retail food prices.
One trend in the results is that a greater proportion of studies are finding no evidence of noncompetitive behaviour in recent years. This could be due to the evolution of empirical methods over
time, as more detailed theories are brought to bear on the problem The recent studies of Paul and
Wohlgenant are in this mould, and they cast doubt on the long held view that the US meat processing
industry is non-competitive. There are other explanations however. One is to state the obvious - that
firms operate in a dynamic environment and they are continually reacting and adjusting, not only to
their competitors, but also to supply and demand changes in the external environment including
changes in the regulatory environment. A greater regulatory presence in the US in the 1990s compared
to the 1980s would surely have some influence on the strategic directions that firms take in industries
which have a “reputation” for non-competitive behaviour.
In the UK in particular, the retailing sector tends to be the chain leader, although that too may be
changing, so most policy attention has been on food retailing. Although the number of published
research studies are few, there has been a recent Office of Fair Trading inquiry and the consensus is
that the major supermarkets do have market power. This matter has been referred to the Competition
Commission. However, there is also evidence that the degree to which the major supermarkets have
been able to exert market power has varied over time as new competitors have entered the market and
as industry conditions have changed. Thus barriers to entry are seen as a very important aspect of an
industry and a very important inhibitor on the ability of firms with market power to maintain noncompetitive behaviour over time.
While very little research has been conducted in Australia using the modern methodologies reviewed
here, this literature does point the way to several potentially useful and related areas of research. The
first is a set of detailed analyses of particular industries of current policy concern using the latest
NEIO techniques, such as the latest work on the US meat processing industry by Paul, and
Wohlgenant. Particularly important here would be to use the newer variable market power indexes,
where the degree of market power, if any, is allowed to vary over time as external influences change.
The second is a set of complementary time series analyses to back up the cost and profit function type
approaches. A start has been made along these lines by Chang and Griffith. Again, the newer methods
allow changing relationships between farm and retail prices at different points in time as external
conditions change.
The third, given that the structure of the food chain in Australia is more similar to the UK than the US,
is a proper analysis of food retail sector performance, like that done for the US by Park and Weliweta.
And the fourth, and perhaps the most interesting given the dynamic aspects of firm behaviour, might
be to have a go at a non cooperative game theory analysis as proposed by Phlips and implemented by
Koontz et al., Slade and others. In such an analysis, the interactions between firms with potential
market power vary over time in a sequence of repeated games. Activity in the Australian retail food
sector would seem to fit very well within such a framework.
One final point to make concerns the influence of trade. Australia is a small country which exports a
significant share of farm output and imports a smaller but still significant share of food requirements.
23
World market prices do matter in the Australian food chain. However most of the models reviewed
here have been constructed in the context of the US and European markets where trade, and the link
between domestic and world prices, is not nearly as important. Any empirical models developed to test
market power in the Australian domestic market have to properly recognise the trade status of the
industry being studied (see Zhao et al).
24
4: Recent Australian Inquiries
During 1999 the results of two public inquiries relevant to the issue of market power in the Australian
food chain were released. These were the findings of the Joint Select Committee on the Retailing
Sector chaired by the Hon. Bruce Baird MP and the Productivity Commission's report the productivity
Commission's report on the effects of National Competition Policy on rural and regional Australia.
Each is discussed in turn but most emphasis is given to the first of them for obvious reasons.
4.1 The Joint Select Committee on the Retailing Sector
The Joint Select Committee on the Retailing Sector which reported in August 1999 (see Australian
Parliament 1999) was asked to inquire into and report on: (a) the degree of industry concentration
within the retailing sector in Australia, with particular reference to the impact of that industry
concentration on the ability of small independent retailers to compete fairly in the retail sector; (b)
overseas developments with respect to this issue, highlighting approaches adopted in OECD
economies; and (c) possible revenue-neutral courses of action by the Federal Government (ie courses
of action that do not involve taxation reform).
Regarding the existence of market power, there is considerable debate in the Report about how to
measure market shares, and such measurements for the three major retail chains range from 80% for
the dry/packaged grocery market (AC Neilsen submission) to 43% for all food and grocery spending
including cafes, restaurants, hotels and taverns (Woolworths submission). Whichever measure is used
however, the evidence is that the share of the major chains is growing over time at the expense of the
independents (ABS commissioned research).
The Committee stated quite categorically that “the market is heavily concentrated and oligopolistic in
nature, where a small number of major chains (Woolworths, Coles and Franklins) each have a
significant degree of economic influence or market power.” (p x).
4.1.1 Winners and Losers
The major winners from this expansion of market share by the major chains were seen to be
consumers, in terms of deregulated trading hours; a greater product choice; lower prices; and the
convenience of one-stop shopping. The Committee stated that “At the consumer level, competition in
the retailing sector appears to be healthy, with retailers vigorously competing with one another on
price and choice. This is evidenced by declining real prices of many grocery items over the last
decade, and a massive expansion in product range to the point where major supermarkets now offer
over 40,000 different items in their larger stores.” (p 2).
The other principal beneficiaries are Australian shareholders (more than 500,000 individuals own
shares in either Coles-Myer or Woolworths), and employees (more than 160,000 individuals across the
three major chains) (p 13). Some suppliers also benefit from increased demand for their products.
The two groups who are losing from this structural change in food retailing are the small independent
competitors, and in many cases, suppliers. The National Association of Retail Grocers of Australia
(NARGA), have been concerned that the major chains have increasingly established themselves
throughout Australia in competition with traditional family-run stores. According to the Report, this
expansion by the major chains has seen the demise of hundreds of small grocery stores, butchers,
bakers, florists, greengrocers, pharmacists, newsagents, liquor outlets and other small retailers as a
result.
Farmers in particular are concerned that the market power of the major chains enables them to drive
very hard bargains in the purchase of produce, which is often done in an aggressive manner. Members
25
of some farm organisations report instances of what they believe to be abuses of market power,
including significant added costs being imposed on suppliers via enhanced labelling and packaging
requirements; the use of various tactics to limit the establishment of brand names by suppliers;
breaches of contract; the "flexible " use of quality standards as grounds for product rejection; the use
of what is said to be exclusive supply agency arrangements in certain markets; and unfair negotiating
practices.
4.1.2 Methods
The evidence used by the Committee to reach its conclusions and make its recommendations was a
combination of factual data on market shares and price changes and largely anecdotal evidence about
firm conduct derived from submissions and hearings. There is some mention in the reported
submissions and in the discussion of developments in other OECD countries. The Committee did not
expound any particular theory used in its deliberations (perhaps understandably, given the status of the
theory) and it seems that no formal empirical studies were undertaken as part of the inquiry.
4.1.3 ACCC Submission
The ACCC submission to the retail sector inquiry provides very useful information about many
matters relevant to our research. It points to the changing nature of relationships in the food marketing
chain, emphasising the increased use of contractual arrangements, and the development of exclusive
contracts with producers in particular.
It warns how attempts by supermarkets and
manufacturers/processors to earn higher returns can impact adversely on farmers and notes that the
sale of raw farm products is now more competitive because of the demise of marketing boards. In
relation to whether in fact the supermarket chains have market power, the submission states as follows
(ACCC 1999, p. 37):
One preliminary issue is whether in fact it can be said that the chains indeed have
substantial market power. While collectively they are clearly a significant voice,
individually none of the chains has more than 35 per cent of the market for warehouse
withdrawals. A firm's market power is related to the structural or behavioural conditions of
a market. Whether a firm has substantial market power in any given case will depend on
the circumstances.
The ACCC goes on to mention a few issues that it claims justify its close watch over the grocery
sector. For example, it is stated that (ACCC 1999, p. 37):
An oligopolistic market structure at the wholesale/retail level of the grocery industry
imposes backward pressure on the agricultural and manufacturing sector which depends on
the chains for the majority of their sales. This causes profits to be squeezed at the producer
level and, to the extent that it drives otherwise viable and competitive players out of the
business, results in a misallocation of resources.
While the ACCC points to various dangers of increased concentration in food retailing, it also points
to some of the benefits from growth of the supermarket chains such as the ability to cater for
consumers with varying income levels, new product development and the convenience of one-stop
shopping.
4.1.4 Recommendations
There are 10 formal recommendations made by the Committee. In these recommendations, “…the
Committee does not seek to invoke protectionist measures for small independent retailers. Rather, it
provides for measures which it believes will enhance competition in the market place.” (p vii).
26
The National Association of Retail Grocers of Australia (NARGA) called for the market share of each
major chain to be capped at 25 per cent, with divestiture taking place within 5 years where any one
chain exceeds that figure. The Committee heard compelling evidence that a market cap would be
unworkable, and would effectively regulate the consumer, and did not recommend such a policy.
A significant body of evidence alleged instances of predatory pricing, where it was said that the major
chains were prepared to lose money indefinitely in certain stores to wipe out the competition. While
the major chains denied these claims, the Committee thought that the evidence was consistent and
widespread, with the common complaint being that the difficulties lie in establishing predatory
conduct under the current provisions of the Trade Practices Act. The Committee believed that the
evidence clearly reveals a need to address the issue of predatory pricing, with a recommendation that
the ACCC be given wider powers to bring representative actions, and to seek damages on behalf of
third parties under Part IV of the Trade Practices Act.
Many complaints received during the course of the inquiry did not raise Trade Practices Act issues,
therefore the Committee saw the need to establish a mechanism outside the ACCC through which
retail industry participants can bring complaints or queries for speedy resolution. The Committee
believed that an appropriate dispute resolution mechanism should take the form of an independent
Ombudsman, to be funded by government, who could attempt to resolve all sorts of complaints
brought to it by businesses in the retailing sector.
The Committee recommended the establishment of a Retail Industry Ombudsman who would be
backed by a mandatory Code of Conduct to regulate conduct in vertically integrated relationships
throughout the supply chain. Being mandatory, the Code of Conduct would enable the courts to take
into account provisions of the Code in determining whether or not business conduct has been
unlawful.
It was also recommended that the Committee should be reconstituted in three years time to re-examine
the retail sector.
4.1.5 Other Comments
•
The Report takes a strong structural perspective in deciding on the existence of market power. For
example, paragraph 2.16 states in part “The Australian grocery retailing industry is oligopolistic in
nature. That is, the market structure is characterised by a small number of firms, each of which
possesses a significant degree of economic influence or market power.” Conversely, most of the
evidence given relates to the conduct of the chains in particular regions or at particular times.
There is no discussion of whether this conduct would eventuate with a different market structure.
•
The Report essentially presents a snapshot of the retail sector during the last few years, with a
little trend analysis. While the data reported in Chapter 2 shows how the retailing sector has
adjusted and responded to external pressures over the years, and Chapter 8 recognises the ongoing
evolution of the retail sector, there is little analysis of the implications of this capacity to change.
Some recent developments are discussed in Appendix A.
•
The Report recognises that the growth of the chains has led to significant economies of size and
scope and that these savings have been, at least in part, passed onto consumers in the form of
lower prices. An implication of this is that market power on the selling side is not such a big
issue.
•
The Report recognises that the market power of the chains on the buying side is enhanced by their
vertically integrated structures, and this leads to some discussion of problems with contractual
arrangements with suppliers, including agricultural producers.
27
•
Finally, the Report mentions but does not evaluate data on profitability provided by the chains.
For example, Woolworths in their submission reported “an EBIT of 3.5% which is amongst the
lowest in the world and is similar to our major competitors.” Coles reported an EBIT of 3.4%. In
some other OECD countries, such data are the primary evidence to decide on whether a
competition inquiry is worthwhile.
4.2 Review of National Competition Policy (NCP) on Rural/Regional
Australia
The Productivity Commission's report on the effects of National Competition Policy on rural and
regional Australia discusses two broad areas of concern directly related to this study. First, it reports
concerns on the part of farmers that the dismantling of the marketing board system that is currently
underway is undermining the ability of farmers to exert countervailing power in their dealings with
large corporations. But the PC did not view this concern as a reason for slowing down reform in
agricultural marketing. It points out that concentration does not necessarily imply lack of competition
in purchasing agricultural products.
Second, it reports concerns about the expansion of retail chains in country Australia; in particular, that
this could result in the demise of smaller players. But the PC states that retailing is highly competitive
and quotes from the Access Economics submission to point to low entry barriers. Moreover, it states
that the former Industry Commission found that profit levels for small and medium food retailing
enterprises in 1993-94 and 1994-95 were higher than for larger enterprises and this was inconsistent
with the notion that large food retailers exercise significant market power.
The PC did not make any specific recommendations with respect to market power issues, viewing
such issues as being the responsibility of the ACCC. However, it does have a chapter on the topic
"National Competition Policy and the Marketing of Rural Products" that is particularly instructive
about the legislative reform process now underway in relation to marketing boards. Among other
things it emphasises the need for case-by-case assessment of the benefits and costs associated with
marketing boards as is occurring under national competition policy.
28
5: The Research Agenda
The primary implication of the author's investigations reported in earlier chapters is that future
research into market power issues in the Australian food market must initially be intensive, case
studies of the marketing chains of major food products or of food products where there is current or
potential policy concern. These will allow development of institutional detail sufficient to reveal the
nature and significance of market power in those chains. The symbiotic relationship between
competitive processes in a market and the market power which flows from success in that competitive
process can be determined when that knowledge is obtained.
The suggested agenda follows from this premise and involves answering a number of major research
questions and sub-questions.
5.1 How Does One Confirm the Existence of Market Power?
While it is very difficult to measure market power in general, it is possible to examine specific
markets with their technologies, for evidence of market power. There are a few things to look for.
5.1.1 Is There Evidence of Long-Run Economic Profits?
The persistence of profits over time is what characterises market power. Potential methods include the
new NEIO techniques, such as the latest work on the US meat processing industry by Paul (1999a,b)
and Muth and Wohlgenant (1999a,b). These methods incorporate variable market power indexes,
where the degree of market power, if any, is allowed to vary over time as external influences change.
Alternate methods include the new time series models, and models of non-cooperative games. The
choice will depend on the chain characteristics and on the available data. The preliminary evidence
presented in Appendix D of the potential of the grains and oilseeds processing/marketing sector to
achieve persistent supernormal profits needs to be confirmed by more detailed case studies. (Further
evidence that this sector is worthy of more detailed examination is seen from two recent ACCC
actions against George Weston Foods Ltd, where this bread and biscuit manufacturer was found to
have entered or attempted to have entered into price-fixing arrangements with retailers.)
5.1.2 Profit Sources
Can one distinguish profits resulting from market turbulence and innovation from market power-based
profits? This is a fundamental subset of the above. It is essential to be able to distinguish efficiency
from the profits of market power, and one of the most obvious efficiency profits is the profit of
technological change and market selection of the efficient from the inefficient. It was this distinction
which the Chicago economists of the 1960s made, and accused the SCP practitioners of ignoring. The
recent work of Paul (1999a,b) attempts to make such a distinction.
5.1.3 Profit Nodes
Can one identify the nodes in the chain at which available profits are taken? Perhaps these could be
called “power nodes” in the chain. The major retail chains could be the nodes here, but there may be
instances where the national manufacturers may enjoy these profits. The differences between the
marketing chains in the US and most other countries might be of great interest to a worker
investigating this phenomenon. Why have national chains not emerged previously in the US? Why
have retail chains not developed “own labels” as a strategic marketing tool in the US while they have
in virtually all other comparable markets? The ability to identify particular nodes will depend
substantially on the data available for that chain. Innovative collections of firm-level data and
“industry” information (from trade magazines and the like) may be required here.
29
One point to make regarding measurement of market power concerns the influence of trade. Australia
is a small country which exports a significant share of farm output and imports a smaller but still
significant share of food requirements. World market prices do matter in the Australian food chain.
However most of the models reviewed have been constructed in the context of the US and European
markets where trade, and the link between domestic and world prices, is not nearly as important. Any
empirical models developed to test market power in the Australian domestic market have to properly
recognise the trade status of the industry being studied.
5.2 What Causes Market Power?
Here one needs to account for inter-firm differences in market performance in the food marketing
chain, using the concepts of resource advantages. Specific issues follow.
5.2.1 Barriers to Entry to Retailing and Processing
Are the sunk costs of intensive marketing efforts such as advertising a barrier to all entry, or are
foreign investors able to leap these hurdles due to easy access to finance? Are there barriers to
upstream entry where the chain has developed innovative institutions? For example, are there barriers
to becoming a franchisee in chicken meat and eggs? Can the non-franchisee still survive in such an
industry?
5.2.2 Importance of Scale of Operation
Can "small" ever be "beautiful" downstream in the food marketing chain? What are the cost
disadvantages of small scale or of non-chain retailing.
5.2.3 What are the Market Niches in the IT Age?
Does efficient consumer response open a niche for small scale enterprise? For example, fresh fruit and
vegetable distribution is changing rapidly: issues such as market bypass by niche retailers as well as
by major chains, direct sales by growers using email or www ordering, etc. and direct purchases by
consumers by www ordering (greengrocer.com) are of interest.
5.2.4 Can Alternative Retailing Systems Co-exist?
Does international and Australian evidence suggest that different techniques of large scale retailers
and of suppliers of independent retailers can survive indefinitely?
5.2.5 Do House Brands Matter?
Why are own brands, or retailer-controlled product marketing, relatively insignificant in the US but
very important in the UK and Australia? Is it due to geography, population size, diseconomies of scale
to retailing, or the US Robinson Patman Act since the 1930s?
5.2.6 Do Regulations Matter?
To what extent is the persistence of long-run market power in the Australian food chain encouraged or
discouraged by government regulation, either domestically or abroad? For example, how do Foreign
Investment Review Board guidelines and rulings influence the structure of the food chain? What is the
effect of import barriers erected by our trading partners on market power expression in the domestic
market? Does one need to model export demand as a separate institutional demand from those of the
major retail chains or the local markets? Again, the institutions by which export demand is expressed
may be important, in terms of where they enter the market (upstream or downstream, or within). The
markets for horticulture exports may be a useful study here.
30
5.2.7 Demand Elasticity
How responsive is retail food demand to changes in prices? The availability of scanner data in the US
has allowed for much improved estimates of the retail demand for food items. Such studies ought now
be possible in Australia. The degree of responsiveness of demand to price changes is, of course, a
traditional indicator of market power.
5.3 Is Contracting an Effective Response to Market Power?
Here the suggestion is to investigate the extent of the use of contracts in Australian farming; the nature
of those contracts; and the desirable features for contracts given the particular circumstances of
Australian agriculture (such as a relatively high level of climatic risk) relative to overseas contract
design experience.
5.3.1 Case Studies
Case studies of primary production in which contracts are a significant issue are of interest. These
could be long established or emerging, and could include horticultural enterprises spanning a range of
product characteristics, from the shortest shelf-life (strawberries?) to durable commodities
(pumpkins?) and seasonal and non-seasonal commodities.
5.3.2 Contracts
Contracts are well established in poultry and dairying and are emerging in beef and other enterprises.
Are agency problems and information asymmetries important? How should contracts be designed? Is
there a role for grower co-operatives? Is there a case for a Code of Conduct (Part IVB of the Trade
Practices Act)? Should farmers be educated for bargaining and other management problems in a
contract supplier role? What can be learned from the US and other countries about all this?
5.3.3 Alternative Institutional Relationships
Are there viable alternative institutional relationships? As responses to agency and information
problems, why have some cases seen vertical integration, others franchising and others the outsourcing
of specific production tasks (as in the chicken meat industry)? So-called "bargaining co-operatives"
have a long history of negotiating price and non-price terms of trade in US agriculture. They are
particularly prominent in the fruits and vegetable industries in California but their presence is much
wider than this. The most common items that are bargained over are price for the farm product, time
and method of payment and quality standards. Other items include methods of grading, premiums and
discounts, production rights and responsibilities and raw product handling procedures to name just a
few (see Iskow and Sexton 1992 for further details). Why have producer co-operatives for
downstream processing survived or recently emerged in some Australian industries and disappeared in
others (and why some are privatising)? What might be the role of the Trade Practices Act and the
taxation system in relation to such co-operatives here? What is the role of new technologies in these
institutional developments? What is the role of opportunistic behaviour by those who first saw
opportunities dealt out by technological and financial developments in the food marketing chain?
5.3.4 Supply Elasticity
What role does supply responsiveness play in a buyers' ability to force down farm-level prices? There
would seem to be scope for more studies of supply response and one focussing on vegetable crops
would be a suitable starting point given that farmer-supermarket relationships in relation to vegetables
has received attention in recent inquiries. It would be important for such a study, which would
probably employ econometric analysis of time series data, to take account of the possibility of a
changing degree of supply responsiveness over time. Not only price response, but also the response of
production systems to demands for quality benchmarking of outputs has to be understood.
31
5.4 Views at the Workshop
It would be fair to say that none of the agenda items suggested at the Workshop drew substantial
criticism as ideas for future research. Rather, those attending gave different levels of support to
different items depending on the perspective they represented.
A number of agenda items are clearly "econometric" in nature. Most Workshop attendees,
understandably, had little or no background in this type of research methodology and, hence, these
items tended to be discussed less than others. Nevertheless, the fact remains that econometric methods
have been, and continue to be, widely used in the US in investigating questions relating to market
power in the food chain. One of the problems that confront researchers in using these methods in
Australia is paucity of data and this received some attention in discussion. One thing that was
encouraging was discussion on the possibility of gaining access to scanner data for use in econometric
analysis. The "flavour" of that discussion was that those possessing the data might make it available
for certain uses.
There was some commentary to the effect that the agenda items relating to the role of contracts
reflected a lack of understanding of "the real world" on the part of the authors. In this final report the
authors have changed the emphasis a little compared to what was originally proposed in relation to
research on contracts. In particular, case studies are now recognised as an essential component of
research in this area.. However, most of the original suggestions remain and the authors believe this is
justified given the fact that there continues to be publicity about abuse of contract conditions on the
part of supermarkets. The suggestions regarding research on alternative institutional arrangements are
closely related to the role of contracts.
One thing that was clear from the Workshop was that future research needed to be useful to decision
makers (a point stressed to the authors by Steering Committee member, Dr Rhonda Smith). In
particular, research that resulted in better ways to gauge whether particular market activities were,
prima facia, anti-competitive was needed. The authors believe that the type of research reported in
Appendix D has much potential in this regard.
32
6. Appendices
6.1 Appendix A: Changes in Food-Chain Management Around the
World
In this Appendix some changes that are occurring in food marketing chains in the US, Europe and
Australia are reviewed.
A.1
Recent Developments in the US Food Market
The discussion here draws heavily on a review paper by Kinsey et al. (1996). The changes that they
describe are argued to be the result of two main causes: changes in consumer preferences and
behaviour, and developments in information technology that have allowed for improvements in the
way the food chain is managed.
A.1.1 Consumer Preferences and Behaviour
In the US, major changes have been occurring in “demographics and lifestyles” (the term used by
Kinsey et al.). In recent years there has been slower population growth compared with, say, 50 years
ago; there is now greater ethnic diversity within the population; the population is aging; and there is
now greater participation of women in the workforce. A development not mentioned by Kinsey et al.
but which is no doubt important is that the workforce spends increasingly less time in heavy, manual
work and more time in semi-skilled work. There has also been slower growth in income and greater
disparity in income distribution.
An implication of these changes, according to Kinsey et al., is that total food sales in the US will not
increase very much in the near future, there will be continued diversification of the consumption
bundle (more ethnic foods) and there will be a greater emphasis on “healthy” food. There will be
increasing emphasis on convenience food which can be taken to include ready-to-heat food purchased
in retail stores, take-away foods purchased in specialist take-away outlets and restaurant meals. In the
US consumers on average are now spending half of their total food dollars on food consumed away
from home. However, greater disparity in income distribution results in two categories of consumers
according to Kinsey et al.: lower-income earners are “economisers and price conscious” in their food
purchase decisions and higher-income earners are “convenience oriented” in their purchase decisions.
These two groups account 45 per cent and 55 percent of consumers, respectively.
In the US the “leadership” of the food chain has always been considered to be the processors and
manufacturers, made up of larger firms and more concentrated industry sectors. The retail sector has
been much less concentrated and more a “follower” of industry trends (Cotteril 1997). These
developments noted above have had some major effects on food retailing. One important effect is
further competition faced by the traditional supermarkets. The economisers are being attracted to socalled “supercentres” such as Wal-Mart that might devote nearly half of their floor space to food items
and the remainder to general discount merchandise. The supercentres have lower operating costs than
traditional supermarkets (17.5 per cent of sales revenue vs. 21.8 per cent). The convenience shoppers
are being attracted to specialist “home meal replacement” providers such as Boston Market (898 stores
in 1996; a total of 3000 planned to be operating in 2000).
The “efficient consumer response” or ECR (see following section) initiative is partly an attempt to put
traditional supermarkets on a par with the supercentres in terms of efficiency. Within traditional
supermarkets there is a trend towards inclusion of take-away counters that offer items such as
barbequed chicken and prepared salads. This is a way of recapturing consumers from the specialist
home meal replacement providers.
33
A preference for pre-prepared food is only one dimension of the behaviour of convenience shoppers
according to Kinsey et al. They also like to maximise the number of tasks that can be accomplished in
one shopping trip or single store visit and, relatedly, they like “ease” of shopping and a minimum of
time to be spent in the task. One response to these preferences has been that supermarkets are now
making non-food items available to consumers either within the supermarket or in adjoining shops that
are leased to specialist providers. The ultimate in convenience, according to Kinsey et al., is the
ability to shop from home using electronic communication.
An important development affecting efficiency is the use of scanner data in re-ordering and its
potential use in payment of firms who supply supermarkets. Scanning data are also used to support
the so-called frequent shopper or loyalty programs increasingly being offered by food retailers. The
scanning data can be used for demographic analysis of customers and analysis of customer preferences
and this information can be used to target products to consumers.
According to Kinsey et al., the food system is now consumer-driven rather than producer-driven.
Once it was the case that farmers made decisions about what food items to produce and in what
quantities, and the resulting production was “pushed through the system from farmer to processor”
(p.14) and so-on down the marketing chain. Volume discounts lubricated the flow of product. In
contrast, the current situation is one in which “power in the system is at the retail end because retailers
receive the information about consumers’ preferences first. This information gives them power to
compete with other retailers, to negotiate with vendors and to respond to consumers.” The information
is shared with agents further up the chain (manufacturers and distributors) and, contrary to what has
traditionally been the case, alliances now exist between these parties. The aim is to respond to
consumers preferences quickly and this implies smaller inventories and distribution costs. More
discerning consumers gives rise to niche markets that can be supplied by small-scale manufacturers:
goats cheese, organic food and micro-brewed beers are examples.
Firms involved in the food
marketing chain now pursue the objective of profit maximisation rather than revenue maximisation.
This commonly made assertion is consistent with the food chain being subject to strong competitive
pressures at the product and firm level, but as significantly, at the level of the capital market. The firm
which does not maximise its profits will face capital market sanctions, and ultimately takeover or
displacement of incumbent management.
An important consequence of the need to respond to more discerning consumers is the production of
raw materials to strict specification, one of the features of a so-called “industrialised agriculture”.
Increasingly farmers are producing under contractual arrangements which embody tightly-specified
product characteristics (e.g., size and moisture content for potatoes). According to Kinsey et al.,
“Contract farming provides secure markets and decreases price risk faced by producers. However,
some farmers may face a virtual monopsony situation in which there is a real imbalance of power.
There will be new incentives for farmers to form cooperatives. A striking example is the North
Dakota wheat producers with their own pasta plant.”
A.1.2 Information Technology
In this section attention is focussed on the ECR initiative. It is described by Kinsey et al. (p. 19) as
“the most comprehensive re-engineering effort [in the food marketing chain]….This industry-wide,
collaborative effort is bringing together food manufacturers, distributors, brokers and retailers. The
aim is to increase both intra- and inter-firm efficiency through new forms of cooperation and
coordination that are based on applications of information technology.” Particularly important are the
opportunities for efficiency gains made available by scanning data collected at retail check-outs. In a
nutshell, advances in information technology have: (a) transformed the ordering process; (b) allowed
access by all firms in the food marketing chain to detailed and timely data on product movement
which, among other things, helps in production scheduling, inventory control and tracking of
consumer preferences; (c) lowered setup and changeover costs on food production lines (shorter
production runs of a greater variety of products are now feasible); and (d) allowed improvements in
warehousing and logistics resulting in lower supply costs and shorter time between receipt of orders
34
and delivery. Selection of product, product replenishment, product promotions and new product
introductions are the areas subject to the ECR initiative. Specific details are provided in Kinsey et al.,
pp 25-33.
One important lesson from the ECR initiative is that it can benefit firms of varying sizes. At least
according to Kinsey et al. (p.34), it can “slow down trends toward increased concentration and vertical
integration”.
A.2
Recent Developments in the European Food Market
While the changes in demographics and lifestyles mentioned above are also occurring in Europe,
increasing concentration in the food marketing chain has become a major policy concern for the last
decade or so. Europe's top 10 grocers had a market share of some 36% in 1997, up from 28% in 1992.
In the UK, the top five grocers had 64% of the market in 1996 (Anon 1999). In addition, it is believed
that retail margins are currently rising and that European food prices are much higher than in the US.
The dependence of such views on general economic and industry conditions is obvious however as
only a year previously commentators were talking about thin margins and sluggish sales in the
European grocery industry (Anon 1998).
One aspect of the concern has been the channel leadership of retailers in the food manufacturing and
distribution system in Europe, particularly in the UK, in comparison to the channel leadership of
manufacturers and processors in the US food chain (Cotteril 1997, Wrigley 1997). The role of retail
proprietary brands is said to be a key issue in this structural difference.
Duke (1992) analysed the changing structure of food retailing in the early 1990s and found that UK
supermarkets had achieved greater power over suppliers by implementing new technology in grocery
retailing and by using cross-Europe purchasing patterns. However he forecast that significant new,
price-cutting entrants would be attracted by the excess profits and that they would be able to overcome
some of the existing barriers to entry and so create a more complex competitive structure.
These forecasts proved quite accurate in just a few years. In 1993, a number of British and European
discount stores entered the UK grocery market, selling a limited range of packaged products at prices
up to 30% below those of the major supermarkets (Anon 1994). These new entrants quickly took
about 10% of the trade, with expectations of this share doubling by the end of the decade. However
the major supermarkets responded by introducing cut-price private labels; by expanding abroad
through mergers and takeovers; and by expanding into new product lines, both food and non-food.
Thus the market share of the discounters has stalled at just above 10%.
The industry has changed again in the last few years with the transition to the European Economic and
Monetary Union. There has been substantial merger and acquisition activity in the retail sector both
within countries and between countries, and Wal-Mart, the largest US retailer, has bought a chain of
German hypermarkets (Anon 1998). Conversely, several large German food retailers (such as Aldi)
have expanded into France, the UK, the US and now Australia.
This merger and acquisition activity has prompted renewed interest by the regulatory authorities. The
UK Office of Fair Trading commissioned a major research study on the methodology of conducting
competition analyses into the retailing sector (London Economics 1997), and then undertook its own
inquiry into food retailing during 1998/99. Subsequently, the Office has referred the grocery retailing
sector to the Competiton Commission for formal investigation (OFT 1999). The European
Commission recently published a White Paper on competition policy reform which has been described
as revolutionary (Financial Times 12/5/99).
A.3
Recent Developments in the Australian Food Market
Australian Parliament (1999) and ACCC (1999) summarise some of the changes that are taking place
in the Australian food chain and, by and large, these changes mimic what is happening in the US and
35
Europe. For example, the “demographics and lifestyles” changes described by Kinsey et al. in the US
are mirrored in Australia: slower population growth, greater ethnic diversity within the population, an
aging population, and greater participation of women in the workforce. Also the major retailers in
Australia have all adopted the ECR type of initiative in an effort to both reduce transactions costs and
be better placed to meet the changing needs of consumers.
Some other developments not really dealt with in these reports suggest that the food chain in Australia
has a significant capacity for change. For example, the changing globalisation of the world food
market has hit Australian shores over the past year. Global food retailers have or are considering the
Australian market: the German retailer Aldi, already operating across Europe and in the US market,
has bought stores in Sydney (some from Franklins), invested heavily in a large distribution warehouse,
and has announced plans to open 100 stores in NSW metropolitan centres (Western Sydney,
Wollongong and Newcastle) over the next few years (Beeby 2000, Mitchell 2000); the US retailer
Wal-Mart, already operating in the UK, has not ruled out investing in the Australian market and has
been touted as a potential buyer for one of the domestic chains; and the South African retailer Metro
now owns most of Davids (and thus has access to hundreds of independents under the IGA, Foodland
etc banners) and has announced aggressive expansion plans. Other major developments are Caltex
Australia planning on opening 200 convenience stores and proper supermarkets, and the expansion of
other convenience store groups.
Finally, a change which is still small in terms of volumes but which has the potential to significantly
change the food marketing chain is the growth of on-line food shopping. New entrants such as
greengrocer.com are gaining sales and the major retailers have responded. Woolworths have
Homeshop trialling in Sydney and Coles Myer have Coles Online in Melbourne. Coles Myer recently
announced the establishment of a new internet division, colesmyer.com, which will include extended
on-line supermarket shopping, and a partnership in an international retailing e-commerce alliance.
Small food retail player David Jones is also developing an internet retailing strategy to launch its
upmarket grocery stores.
36
6.2 Appendix B: A Model of Marketing Margins Under Perfectly
Competitive and Pure Monopoly/Monopsony Behaviour
The development of the model draws heavily on Gardner’s 1975
paper. Indeed, the model representation in equation form is
identical to that outlined by Gardner with two exceptions. First, a
more general representation of optimal input use is included here,
namely, employ each input at a level such that marginal revenue
product (marginal product times marginal revenue) is equal to
marginal input cost. (Gardner assumes perfect competition in the
output and input markets such that marginal revenue equals output
price and marginal input cost equals the input price. Hence, the
optimisation condition requires equating input prices to the product
of marginal product and output price). Second, there are some
minor differences in notation.
Two inputs are used in the food marketing industry, a farmproduced input (e.g., wheat; "farm input" for short) and a
"conglomerate" marketing input (transport, storage, milling,
packaging etc.) to produce a retail product (e.g., flour) purchased
by final consumers. Interest centres on the behaviour of the
marketing margin (the relationship between the output price and
the price of the farm input) in response to changes in exogenous
influences.
B.1
Notation
x = quantity of retail product;
a = quantity of farm input;
b = quantity of the marketing input;
f i = marginal product of input i;
f ij = partial derivative of f i with respect to input j;
Pi = price of quantity variable i;
N = exogenous shifter of retail demand;
W = exogenous shifter of supply of farm input;
T = exogenous shifter of supply of the marketing input;
η = price elasticity of retail demand;
37
η N =elasticity of retail demand with respect to N ;
ei = price elasticity of supply of input i ;
eT = elasticity of supply price of the marketing input with respect to T ;
eW = elasticity of supply price of the farm input with respect to W ;
S i = share input i ( Pi i / Px x );
σ = elasticity of substitution between inputs a and b in producing x ;
χ i = elasticity of production of input i ;
APi = average product of input i ;
MFC i = marginal input cost of input i ;
MRPi = marginal revenue product of input i ;
β 1 = (1 + (1 / η )) /(1 + 1 / ea )) ;
β 2 = (1 + (1 / η )) /(1 + 1 / eb )) ; and
Ei , j = total elasticity of variable i with respect to variable j .
B.2
The Model
(B.1)
x = f ( a, b)
(production function)
(B.2)
x = D ( Px , N )
(retail demand function)
(B.3)
Pa = β 1 Px f a
(optimal use of input a )
(B.4)
Pb = β 2 Px f b
(optimal use of input b )
(B.5)
Pa = h(a,W )
(supply input a )
(B.6)
Pb = g (b, T )
(supply input b )
The endogenous variables in the system are the prices and quantities of the two inputs and the retail
price and quantity (six in total). It is assumed that the production function exhibits constant returns to
scale and, hence, the following relationships hold:
(B.7)
f ab = f ba =
fa fb
σx
38
(B.8)
f aa = −
bf a f b
aσx
(B.9)
f bb = −
af a f b
bσ x
Other useful relationships are:
(B.10) χ i = f i / APi
(B.11) g a =
Pa
ea a
(B.12) g b =
Pb
eb b
(B.13)
fa =
Pa
β 1 Px
(B.14)
fb =
Pb
β 2 Px
(B.15) S a =
Pa a β 1 Px f a a
=
= β1 χ a
Px x
Px x
(B.16) S b =
Pb b β 2 Px f b b
=
= β2χb
Px x
Px x
(B.17)
Sa Sb
+
= χ a + χ b = 1 under CRS
β1 β 2
(B.18) S a β 2 + S b β 1 = β 1 β 2 (from (17))
B.3
Constraints on Betas
Under the assumption that firms are price takers in both the output and input markets, β 1 and β 2
equal 1.0, since the elasticities of retail demand and input supply will be viewed as infinite. Here it is
assumed that firms always act as price takers in purchases of input b and, hence, the denominator of
β 2 is always 1.0. In the case where the firm acts as a monopolist in the sale of x , profit maximising
behaviour requires that it must be operating at a point on the retail demand function where demand is
relatively elastic, assuming that marginal costs are positive. Hence, the numerator of β 1 and β 2 has a
maximum value of 1.0. The denominators of β 1 and β 2 will have a minimum value of 1.0 and will
increase as the elasticities of input supply decrease, with a maximum value approaching infinity. It
can be concluded, therefore, that both β 1 and β 2 will range between 0.0 and 1.0.
39
B.4
Measures of the Marketing Margin
There are various ways the marketing margin can be measured:
•
Price difference or margin = Px − Pa
•
Price ratio = R = Px / Pa
•
Farmer's share of "the food dollar" = S a
•
Percentage margin = % M = ( Px − Pa ) * 100 / Pa = (( Px / Pa ) − 1) * 100
These are all closely related to each other. Attention here focuses on the last three measures. In
addition, attention is given to the elasticity of price transmission associated with a particular
exogenous shifter. Gardner (1975) defined this as the ratio of the percentage change in Px associated
with a change in Pa where either Px or Pa was subject to price control. Here it will be defined, say
for the exogenous shifter N , as:
•
Elasticity of price transmission ( φ )= Eφ , N = E Pa , N / E Px , N
and analogously for the shifters W and T . It is of interest since farmers often complain that changes
in farm-level prices are not reflected in changes in retail prices; in particular, there are often claims
that falls in farm-level prices are not reflected in lower retail prices.
B.5
Deriving Analytical Results
There are various ways of proceeding to derive total elasticities showing the percentage change in one
of the margin measures resulting from a one per cent change in an exogenous shifter. Gardner shows
how the model can be reduced to three equations by equating (B1) and (B2), (B3) and (B5), and (B4)
and (B6). Then, if one wants to study the effect of, say, a shift in retail demand (i.e., a change in N ),
first totally differentiate the three-equation system with respect to N , make substitutions using
various of the relationships defined and then write the resulting three-equation system in matrix form.
The result is:
 
–  Sb + 1
  β 2σ ea

Sa
(B.19) 

β 1σ

Sa

β1





Sb
β 2σ
 S
1
–  a + 
 β 1σ eb 
Sb
β2

1   E a , N  0 

  

  
1   Eb , N  = 0 

  

  
– η   E Px , N  η 
  N


or, using (15) and (16),
40
  χb
1
+
 – 
ea
  σ

χa
(B.20) 
σ

χa






χb
σ
 χ
1
–  a +
eb
 σ
χb




1   E a , N  0 
  


  

1   E b , N  = 0 
  

  
– η 
  E Px , N  η N 

where all parameters are measured at their initial equilibrium level. The solution to the vector of total
elasticities can then be found by Cramer’s Rule or matrix inversion. The total elasticities not included
in the vector above can be derived from the solutions to those that are included.
One can take a slightly different route and solve for a vector of three total price elasticities with
respect to, say, the exogenous shifter N . The resulting matrix equation is:

− S a ea
− S b eb 

η
β1
β2
  E  − η 

P ,N
N

 S a eb
  x  
S a ea
(B.21)  1
− 
+ 1  E Pa , N  =  0 
β 1σ
 β 1σ
  E   0 


  Pb , N  

 − S b ea

S b eb


1
−
+
1


 β σ

β 2σ
 2



or, using (B15) and (B16):


η
− χ a ea
− χ b eb   E  − η 

Px , N
N
χ a ea
 
 χ a eb
 

(B.22) 1
−
+ 1  E Pa , N  =  0 

σ
 σ
 


 E P , N   0 
χ
e
χ
e


b b
 b
 1 −  b a + 1


σ
 σ

Equations (B19) and (B21) give solutions to total elasticities for the three prices and quantities. Total
elasticities for the marketing margin measures are simple functions of the price and quantity total
elasticities. For example, the total elasticities with respect to N are:
(B.23) E R , N = E Px , N − E Pa , N
(B.24) E S a , N = E Pa , N + E a , N − E Px , N − E x, N
(B.25) E % M , N = E R , N R /( R − 1)
Those with respect to W and T are the same except that the N subscript changes to W or T.
From (B20) and (B21) it is apparent that there will be one situation in which the solutions for the total
price and quantity elasticities, the total elasticities for R and S a , and the price transmission elasticities
will be invariant to assumptions about the degree of competition: this will be when the output demand
and input supply functions take the constant elasticity form and when the production function is such
41
that χ a , χ b and σ are constant irrespective of the output level (the Cobb Douglas functional form
satisfies these conditions). In other cases the solutions will depend on the nature of functional forms
because the base parameter values will vary with the nature of the competition in the output and input
markets. For example, if the output demand function is linear, monopoly in the output market would,
ceteris paribus, imply a lower output and a higher price than under perfect competition in the output
market and, hence, a higher own-price elasticity at the initial equilibrium. The elasticity of the
percentage margin (%M) is dependent on the base levels of Px and Pa and, hence, will always differ
according to the nature of competition assumed.
B.6
The Formulae
The formulae for the various total elasticities of interest are provided in Table B.1. Further details on
the derivations are available in the appendix to the Gardner (1975) paper and from the authors.
Table B.1: Total Elasticity Formulae*
Elasticity
1% increase in N
1% increase in W
1% increase in T
Price ratio
E R, N =
E R ,W =
E R ,T =
β 1η N S b (e a − eb ) / D
β 1 e w e a S b (η − eb ) / D
β 1 eT eb S b (e a − η ) / D
E Sa , N =
E Sa,W =
E Sa,T =
(R)
Farmers'
share
β 1η N S b (e a − e b )(σ − 1) / D β 1 eW e a S b (η − eb )(σ − 1) / D
β 1 eT eb S b (e a − 1)(σ − 1) / D
(Sa)
Percentage
margin
E%M , N =
E R , N R /(R − 1)
E % M ,W =
E % M ,T =
E R ,W R /( R − 1)
E R,T R /( R − 1)
(%M)
Price transmission
(φ)
Eφ , N =
Eφ ,W =
Sa
S
eb + b ea + σ
β1
β2
eb + σ
β 1 (e b − η
Sb
) + S aσ
β2
S a (e b + σ )
E φ, T =
η+ σ
ea + σ
* D = −η ( β1 S b e a + β 2 S a eb ) + β 1 β 2 e a eb + σ ( β1 S b eb + β 2 S a e a )
B.6
Numerical Example
An example of equations (B1) through (B6) in which the retail demand and input supply functions are
of the constant elasticity form and the production function has a constant elasticity of substitution is as
follows:
ρ
ρ
(B.26) x = A1 (a + b )
1
ρ
η
(B.27) x = A2 Px N η N
ρ
(B.28) Pa = β 1 Px A1 ( x / a )1− ρ
42
ρ
(B.29) Pb = β 2 Px A1 ( x / b)1− ρ
(B.30) Pa = A4 a 1 / ea W eW
(B.31) Pb = A3 b1 / eb T eT
The following parameter values are assumed:
ρ = −1.0;η = −1.5;η N = 1.0; e a = 1.0; eb = 2.0; eW = 1.0; eT = 1.0
A1 = 1.0; A2 = 1000.0; A3 = 0.00125; A4 = 0.017678
N = 100.0;W = 100.0; T = 100.0
The implied value of σ (defined as 1 /(1 − ρ ) ) is 0.5.
Values for the β 1 and β 2 are also implied by the assumed parameter values. These are summarised in
Table B.2.
Table B.2: Implied Values for β 1 and β 2 *
Purchasing of a :
Sale of x :
Perfect competition
β1
Monopsony
β2
β1
β2
Perfect competition
1.0
1.0
0.5
1.0
Monopoly
0.33333
0.33333
0.16667
0.33333
*Assumes purchasing of b is perfectly competitive.
The base solutions for the model (i.e., when all parameters are set at the assumed or implied values)
are shown in Table B.3.
Table B.3: Base Solutions
Variable
Nature of competition *:
C-C
M-C
C-M
M-M
x
100.000
45.240
78.916
36.230
a
200.000
97.474
142.883
70.093
b
200.000
84.423
176.272
74.993
Px
100.000
169.688
117.100
196.767
Pa
25.000
12.184
17.860
8.762
Pb
25.000
16.243
23.470
15.309
margin
75.000
157.503
99.240
188.006
300.000
1292.679
555.643
2145.770
% margin
Px/Pa
4.000
13.927
43
6.556
22.458
Sa
0.500
0.155
0.276
0.086
Sb
0.500
0.179
0.448
0.161
*C-C denotes perfect competition in the sale of x and perfect competition in the purchasing of a; MC denotes monopoly in the sale of x and perfect competition in the purchasing of a; C-M denotes
perfect competition in the sale of x and monopsony in the purchasing of a; and M-M denotes
monopoly in the sale of x and monopsony in the purchasing of a.
The total elasticities associated with changes in the exogenous shifters are shown in Table B.4.
Table B.4: Total Elasticities
Elasticity of:
Nature of competition*:
C-C
M-C
C-M
M-C
with respect to N:
Px/Pa
-0.087
-0.094
-0.077
-0.084
% margin
-0.116
-0.101
-0.091
-0.088
Sa
0.043
0.047
0.039
0.042
Price trans.
1.250
1.273
1.218
1.240
with respect to W:
Px/Pa
-0.304
-0.328
-0.270
-0.293
% margin
-0.406
-0.354
-0.319
-0.307
Sa
0.152
0.164
0.135
0.147
Price trans.
2.400
2.616
2.135
2.309
with respect to T:
Px/Pa
0.435
0.469
0.386
0.419
% margin
0.580
0.505
0.455
0.438
Sa
-0.217
-0.234
-0.193
-0.209
Price trans.
-0.667
-0.667
-0.667
-0.667
*C-C denotes perfect competition in the sale of x and perfect competition in the purchasing of a; MC denotes monopoly in the sale of x and perfect competition in the purchasing of a; C-M denotes
perfect competition in the sale of x and monopsony in the purchasing of a; and M-M denotes
monopoly in the sale of x and monopsony in the purchasing of a.
B.7
Qualitative Comments
The first thing to note is that, under conditions of perfect competition in the sale of the retail product
and in the purchasing of input a (purchasing of input b is assumed to be perfectly competitive in all
scenarios), the values of β 1 and β 2 are both 1.0 and the formulae shown in table B.3 reduce to those
reported in Gardner (1975) in which perfect competition in all three markets was assumed.
The second thing to note is that, because β 1 and β 2 are both positive, the signs of the total elasticities
are invariant to the nature of the competition assumed. This means that, in general, the sorts of
44
qualitative statements made by Gardner about margin behaviour under perfect competition apply also
to the other forms of competition. This is not surprising because the various forces that govern
adjustment to exogenous changes under perfect competition (e.g., substitution between inputs in
response to relative price changes, with the degree of substitution depending on σ ) also apply in the
case of other forms of competition. Thus, for example, even if there is monopoly in the sale of the
retail product and monopsony in the purchase of the farm input, an increase in demand for the retail
product will increase the farmer's share of the food dollar provided the farm input is more elastic in
supply than the marketing input and the elasticity of input substitution is greater than unity, or if the
farm input is less elastic in supply than the marketing input and the elasticity of input substitution is
less than unity
General qualitative statements beyond the two mentioned above, such as whether a particular total
elasticity is greater or smaller under perfect competition than under the monopoly-monopsony
scenario, are not possible because base parameter values may well be different under different types of
competition. One would need to know how parameters change with different forms of competition and
this will depend on underlying functional forms for the demand and supply functions and for the
production function. The exception already mentioned is when all of the functions are of constant
elasticity form in which case the total elasticities, except that for the percentage margin whose value is
dependent on the base margin, will be the same across the different forms of competition.
Some of the important results which follow from the Gardner model (see Gardner 1975, pp. 406-407)
are listed below. The particularly important point that follows from the extension of the Gardner
model reported here is that these important results hold under all the competition scenarios
investigated here.
•
Farm and retail level food prices move together in different ways depending on whether the events
that cause the movement arise from a shift in retail demand, farm supply or the supply of
marketing services.
•
Events that increase the demand for food will reduce the retail-farm price ratio and percentage
marketing margin if marketing inputs are more elastic in supply than farm products and viceversa.
•
Events that increase (decrease) the supply of farm products will increase (decrease) the retail-farm
price ratio and percentage marketing margin.
•
Events that increase (decrease) the supply of marketing inputs will decrease (increase) the retailfarm price ratio and percentage marketing margin.
•
The percentage price spread is analytically distinct from the farmer’s share of the food dollar and
the two will behave differently under changing market conditions unless σ=0. If σ=0, the
farmer’s share is constant. If σ>1, an increase in the marketing margin will be accompanied by an
increase in the farmer’s share of the food dollar. Otherwise, lower marketing margins go together
with an increased farmer’s share.
What these results suggest, among other things, is that it is dangerous to draw inferences about the
behaviour of marketing firms from simple concepts as the farmer’s share of the consumer’s dollar.
Even under conditions of perfect competition the farmer’s share can decrease in the face of certain
types of changes depending on various parameter values. Also, much depends on the crucial
parameter σ that measures the degree of substitution between the farm product and other marketing
inputs in producing the retail product.
In short, what this Appendix has shown is that the direction of movement in the various measures of
the marketing margin in response to exogenous changes in demand and supply are the same
irrespective of which of the extremes of competition investigated here actually hold in practice. The
45
actual values of the marketing margin measures, and the extent of movement in those measures in
response to exogenous demand and supply shifts, are quite a different story. These can be expected to
be generally different depending on which of the extremes of competition actually hold. Outcomes
will depend on underlying functional forms.
46
6.3 Appendix C: Some Comparative Statics on the Incentive for
Collusive Buying Behaviour
Consider the following market model for a raw material (e.g., processing apples) where the total
demand has been partitioned into two sources, a and b:
a
a
b
b
(C1)
D =D (P, X)
(C2)
D =D (P, Y)
(C3)
S=S(P, Z)
(C4)
D +D =S
a
b
Demand is denoted by D, supply by S, price by P and X,Y and Z are shift variables.
a
The aim is to investigate the impact on price if there is a rightward shift in D of a certain proportion
resulting from a change in X. This is done using basic comparative statics; in particular, totally
differentiate the system of equations and transform partial derivatives and differentials to elasticities
and proportionate changes, respectively. After doing this one obtains the following expression for EP,
the proportionate change in price:
(C5)
EP= -xra/[raηa+(1-ra)ηb-ε]
a
where x represents the proportionate rightward shift in D measured from the initial equilibrium, ra
represents the share of demand source "a" in total demand measured at the initial equilibrium,
η denotes demand elasticity and ε denotes supply elasticity.
Clearly, the proportionate change in price is dependent on three elasticity coefficients, market share
and the extent of the rightward shift in demand. If the incentive for collusive buying behaviour is
positively related to the size of the potential price increase, then a large value for ra does not
necessarily imply a strong incentive. In particular, a high value for ra can be offset by a high absolute
value for e, ηa or ηb.
The basic model can be altered in various ways. One alternative is to conjecture that those buyers who
b
make up D hold their purchases of the raw material constant in the face of a demand increase from
b
source "a" (in the analysis above there is an implicit assumption that D adjusts to the higher price
a
b
resulting from the increase in D ). This would amount to an increase in D occurring simultaneously
a
b
with the increase in D . One can solve for the required rightward shift in D that would keep the
quantity purchased by "b" buyers constant in the face of the higher price (which was the conjecture
used by Helmberger and Hoos 1965). Another possibility is to allow for substitutes in demand and
supply (possibilities include a fresh/processed dichotomy or an Australian/imported dichotomy).
Now consider the impacts of a supply shift. The expression for the proportionate change in price is
now given by:
(C6)
EP= z/[raηa+(1-ra)ηb-ε]
where z represents the proportionate horizontal shift in the supply function measured at the initial
equilibrium point (z is negative for a supply decrease).
47
The same general statements about the dependency of EP in the case of a demand shift also apply for a
supply shift. In particular, the degree of buyer concentration is only one of the factors governing the
degree of price rise.
48
6.4 Appendix D: Some New Empirical Results
Given the paucity of empirical evidence on the competitive structure of the food marketing chain in
Australia that is available to assist policy makers, an attempt was made to add to the stock of
knowledge. The objectives in undertaking this research were to take a broad sweep across all sectors
from farm gate to the consumer and across a wide range of 14 food products of varying levels of
processing; and to use a modelling framework which was broadly applicable across these products
using readily and publicly available data. Such a study should provide the sort of filter requested by
ACCC and others, so that detailed case studies on the competitive structure of the marketing chain do
not have to be undertaken on every single food market deregulation or merger proposal.
D.1
Method
A marketing margins framework is adopted as the basis for the study, and of the many alternative
models available (recently reviewed by Wohlgenant 1999), the structural NEIO framework developed
by O’Donnell (1999) is used. The idea is that the observed marketing margin for a food product
potentially contains three components. The first is the costs of providing the marketing services
required to transform the agricultural input into the food product. The second is any rent due to noncompetitive buying behaviour in the relevant input market, due to any divergence between input price
and marginal factor cost, and the third is any rent due to non-competitive selling behaviour in the
output relevant market, due to any divergence between price and marginal revenue.
Following the notation set out by O’Donnell, let pm be the price of the food output m, wm the price of
the agricultural input m, qm the aggregate quantity of output m and xm the aggregate quantity of input
m. An inverse demand function operates in the output market of the form pm = f(qm). Agricultural
marketing firms combine agricultural inputs xm, with input supply functions of the form xm = f(wm) and
non-agricultural inputs z, to produce qm. Making the common assumptions of fixed proportions, and
constant returns to scale (as for example Hyde and Perloff 1998), specific assumptions about
functional forms for the demand, supply and cost functions, and aggregating over all n firms in the
industry, the first order condition from the profit maximisation problem eventually results in an
estimable equation for the marketing margin of the following form (full details of the derivation and
the estimation model are given in O’Donnell 1999):
K
(D1)
mj = aj + ∑ cjkzk + βjqj +
k =1
M
∑γ
x / wm
jm m
m =1
where, for any product j:
mj = pj -wj, is the industry marketing margin;
zj = non-agricultural input prices and trend and seasonal factors if required;
βj = -θqjj/ηj, where ηj is the slope of the market demand function for product j and θqjj is the
conjectural elasticity of the average firm in the output market with respect to aggregate output; and
γjm/wm = θxmj/εjm, where εjm is the slope of the input supply function for agricultural input j and θxmj is
the conjectural elasticity of the average firm in the input market with respect to aggregate inputs, and
where there may be more than one input m contributing to output j.
From theory and the assumptions made, the βj, γjm and cjk coefficients must also be non-negative.
Thus the industry marketing margin for a food product can be expressed as a linear function of the
prices of marketing inputs and two expressions containing the quantity of the agricultural input (or
output). These latter two expressions represent output and input market market power respectively. If
49
output and input markets are competitive, the conjectural elasticities are zero and the margin equation
reduces to the familiar condition of the price of marketing services equals the marginal cost of
supplying them. Thus a test of competitive behaviour in a particular food product market or
agricultural input market is simply a test on whether the βj and γjm coefficents respectively are
positively significantly different from zero. No direct estimates of the conjectural elasticities are
provided if these coefficients are significant, but they can be inferred if estimates are available of the
demand and supply elasticities.
The modelling approach undertaken was to estimate a set of equations of the form given in equation
(1) for the products of interest with the non-negativity restrictions imposed using the nonlinear
regression command in TSP. Single equations and SUR systems were estimated, with and without a
dummy variable (1987 onwards=1) to account for the greater concern with concentrated markets in the
last decade, and using production data as the measure of throughput. The results reported below are
for the SUR product group systems, without dummy variables, and with an aggregate cost index in
place of individual cost variables because of multicollinearity problems. The dummy variables did not
produce results much different from those reported.
D.2
Data
For any one product, the only data required are farm and retail prices, the quantity produced, the costs
of supplying marketing services and the price of all other goods (CPI) for the normalised cost
function. Annual data on these required variables for the 14 food products listed in Table 1, over the
period 1970 to 1997 where possible, were taken from readily available sources as detailed below. All
price variables were converted to real terms so the dependent variables in the regressions are real
margins.
General Data
•
Consumer Price Index, Australia, all groups, base 1980/81, both calendar year and financial year,
both rebased to 1990, ABS 6401.0 and ABARE
•
Population, Australia, million, both calendar year and financial year, ABS and ABARE
•
Wages Index, Australia, all adults weekly, base 1980/81, both calendar year and financial year,
both rebased to 1990, ABS
•
Interest rate, Australia, 90 day bank bills, financial year, calendar year calculated, ABARE
•
Electricity cost index, Australia, base 1990, financial year, calendar year calculated, ABS 6411.0.
•
Marketing Cost Index, calculated as (0.75*Wage)+(0.1*Electricity)+(0.15*Interest), both calendar
and financial year, base 1990
Meat Products Data
•
Retail prices for meat, c/kg, calendar year, ABARE
•
Farm prices for livestock except chicken, c/kg, calendar year, ABARE
•
Farm price for chicken, average unit gross value of livestock slaughterings, poultry, financial year,
$ per bird, converted to c/kg by average carcase weight and to calendar year, ABS 7503.0
•
Production and aggregate domestic consumption, kt carcase weight, calendar year, ABARE and
MLA
50
•
Meat models estimated on a calendar year basis
Grains and Oilseeds Products Data
•
Retail prices, price for various pack sizes converted to c/kg, converted to financial year, ABARE
and ABS 6403.0
•
Farm prices, unit gross value of production, $/t converted to c/kg, financial year, ABARE and
ABS 7503.0
•
Production and aggregate domestic consumption, kt, financial year, ABARE (no reliable
consumption data for rice or maize)
•
Grains and oilseeds models estimated on a financial year basis
Fresh Fruit and Vegetable Data
•
Retail prices, c/kg, financial year, ABARE and ABS 6403.0
•
Farm prices, average unit gross value of principal crops, $/t converted to c/kg, financial year, ABS
7503.0
•
Production, kt, financial year, ABARE and ABS 7503.0
•
Aggregate domestic consumption, apparent per capita consumption, kg/head, financial year,
converted to aggregate consumption, ABARE and ABS Fruits Australia,
•
Fruits and vegetables models estimated on a financial year basis
D.3
Results
The results of estimating equation (1) are quite striking (Table D1). First, there are a large proportion
of corner solutions where the constraints on both the input and output conjecture coefficients were
binding at zero. Thus the null hypothesis of a competitive market in both output and input markets
could not be rejected for any of the meat products, fresh fruits or fresh vegetables.
51
Table D1: Constrained SUR Estimates
Food Product
Cost index
coefficient
Output conjecture
coefficient
Input conjecture
coefficient
Trend
coefficient
0.00**
0.00**
-0.70
Meat Products
Beef
228.5
Lamb
23.41
0.00**
0.00**
-4.98**
Pork
45.56
0.00**
0.00**
-4.31**
0.00**
0.00**
-8.33**
-2.57**
Chicken
0.00**
Grains Products
Rice
63.73*
0.00**
0.00**
Bread
61.62**
0.00**
0.06 (.26)
0.00**
1.70**
0.00**
1.06 (.21)
9.19**
0.00**
0.00**
2.35**
0.00**
0.00**
-1.04
0.00**
0.00**
-0.40
Margarine
Breakfast cereal
230.0**
0.00**
2.03**
-8.27**
Fresh Fruit
Oranges
Bananas
39.04
0.00**
Fresh Vegetables
Potatoes
34.36
Tomatoes
0.00**
0.00**
0.00**
-1.75
Carrots
0.00**
0.00**
0.00**
-0.06
0.00**
0.00**
1.55
Onions
108.9**
* significant at the 10% level; ** significant at the 5% level; one-tail test for the potential
constrained coefficents; two-tailed for the time trend.
The result in relation to these output markets matches the conclusions from the Joint Select Committee
noted elsewhere in this report and the views of the Prices Surveillance Authority (1994), which
regarded the markets for meat and fresh fruit and vegetables as “competitive” (p.14). It also accords
with previous evidence on meat products (Zhao et al 1998; Hyde and Perloff 1998). The result in
relation to input markets is consistent with the evidence for meat products in Chang and Griffith
(1998), but is somewhat contrary to the tenor of submissions to the Joint Select Committee in relation
to the fresh fruit and vegetable sector.
52
In the processed grains and oilseeds sector of the food market, the output conjecture coefficients are
not positively significantly different from zero and thus indicate a competitive consumer market for
the relevant food products, as suggested by the Joint Select Committee. However three of the input
conjecture coefficients are positive, one is highly significant and the other two are significant at just
over the 20 per cent level. This provides some evidence of a non competitive buying market for the
relevant farm commodities. Given the assumptions made, and the fact that the estimated coefficients
reflect average behaviour over a 20 year period, these results suggest that non competitive activity has
been a persistent feature of this market sector.
This result also accords with the views of the Prices Surveillance Authority (1994), which regarded
the markets for products contained in the Breakfast Cereals and Cooking Oils and Fats indexes as “not
effectively competitive” (p.14) and consequently maintained price surveillence on the major firms in
this product group (at the time Arnotts, Kelloggs, Uncle Tobys and Sanitarium).
The high proportion of corner solutions (also found by O’Donnell 1999) requires comment. The
equations were re-estimated without the non-negativity constraints, and for only two of the products,
rice and bread, did the unconstrained equations produce positive signs on all the coefficents required
to be non-negative. It is no surprise that these products are both in the grains and oilseeds processing
sector. In all other cases, at least one of the quantity-related coefficients, and in some cases the cost
index coefficient also, was estimated to be negative and significantly different from zero. Therefore
the estimated equations strongly suggest a negative relationship between margins and throughput. This
implies a declining average cost curve for the processing and distribution sector for these food
products, which calls into question the constant returns to scale assumption. This adds further weight
to the call by Paul (1999a,b) for greater attention to be paid to underlying cost structures when
examining the competitive nature of the food chain. Of course more detailed data sets are required for
this, and a broad coverage of food products as undertaken here may no longer be possible.
Another comment concerns the general significance of the trend coefficients. These trends were
included to account for factors apart from those formally specified thought to have an influence on the
pricing of market services - the increased demand by consumers for additional market services as well
as deregulation and takeover and merger activity. The mix of positive and negative trend coefficients
suggests that there are different balances of these factors across the food markets studied. Again,
knowledge of the market is important.
A final comment is that studies such as reported here, based on average representations of market
behaviour over some long historical period, may not be very useful in identifying episodes of noncompetitive behaviour in a timely manner, when required by regulatory authorities. This general
problem with time series-based models suggests that even more attention should be given to
monitoring and understanding the institutional detail of the market in question, and to attempting to
apply those models having variable market power indexes. If variations in the use of market power can
be related to particular changes in the environment facing the firm or market in the past, such changes
can be taken as leading indicators of possible variations in the use of market power in the present and
in the future.
D.4
Implications
The primary implication of the literature review part of the paper is that future research into market
power issues in the Australian food market should be in the form of intensive case studies. These will
allow development of the institutional detail sufficient to reveal the nature and significance of market
power in those chains and how and why it ebbs and flows over time. However such resource-hungry
analyses cannot be undertaken for every market faced by a merger or deregulation proposal. The
implications of the empirical results provided in this Appendix are that (a) attention should be focused
on the supplier side, not on the consumer side of the market (an implication fully supported in the
Joint Select Committee report); and (b) the sector providing the most likely payoffs from greater
research effort would be the grains and oilseeds processing and distribution sector. Given these
53
implications are likely to continue to be relevant into the near term future at least (Australian
Parliament 1999), a research agenda focussing on these issues would seem justified.
•
The preliminary evidence of the potential to achieve persistent supernormal profits in the grains
and oilseeds marketing sector needs to be confirmed by more detailed case studies. This could be
done by application of the more detailed NEIO techniques, such as the latest work on the US meat
processing industry by Paul (1999a,b) and Muth and Wohlgenant (1999a,b). These methods
incorporate variable market power indexes, where the degree of market power, if any, is allowed
to vary over time as external influences change, and several also allow separation of operational
efficiency gains due to scale and scope and technical innovation. The choice of method will
depend on the chain characteristics and on the available data. Baker and Bresnahan (1992) and
Hyde and Perloff (1995) examine the advantages and disadvantages of various methods. However
innovative collections of firm-level data and “industry” information (from trade magazines and the
like) are likely to be required if detailed case studies are to be completed.
•
An important point to make regarding measurement of market power concerns the influence of
trade. Australia is a small country which exports a significant share of farm output and imports a
smaller but still significant share of food requirements. World market prices do matter in the
Australian food chain, and the export parity price can often be regarded as a floor for products that
have only a minor degree of processing. However most of the formal models reviewed have been
constructed in the context of the US and European markets where trade, and the link between
domestic and world prices, is not nearly as important. Any more detailed empirical models
developed to test market power in the Australian domestic market should properly account for the
trade status of the industry being studied.
•
Another important point concerns the implication from the empirical work of increasing returns to
scale and non-optimal plant sizes in many industries over the sample period. While this situation
seems to be changing rapidly, especially in meat and milk processing, further research needs to
examine in detail the shape of cost curves and returns to scale in these industries (Paul 1999a,b).
54
6.5 Appendix E: The Workshop
A Workshop was held in the Boardroom, NSW Farmers' Association, Sydney on 12 November, 1999
to discuss the suggested research agenda flowing from the project. Participants were issued with a
background paper titled "Market Power in the Australian Food Chain: Towards a Research Agenda".
Some 30 invitations were issued to a wide cross-section of individuals thought to have a key interest in
the research agenda. In the event, 19 invitees were able to participate. A Commissioner from the
ACCC had to withdraw (understandably) from the Workshop at very short notice because of litigation
that the ACCC had underway and it was not possible given the short notice to arrange for a
replacement. The Hon. Bruce Baird who chaired the Joint Select Committee on the Retail Sector
provided a keynote address overviewing the Committee's work and its outcomes.
The Workshop participants were:
Dr Fredoun Ahmadi, University of Sydney
The Hon. Bruce Baird, MP
Ms Jayne Baric, Australian Horticultural Corporation
Dr Peter Barnard, Meat and Livestock Australia
Mr Jim Booth, NSW Cabinet Office
Dr Tony Dunne, Gatton College, University of Queensland
Dr Jeff Davis, Rural Industries Research and Development Corporation
Dr Craig Freedman, Macquarie University
Dr Garry Griffith, NSW Agriculture and University of New England
Ms Andrea Griffiths, The Senate
Mr Michael Keogh, NSW Farmers' Association
Mr Michael Kovas, Food Retailers' Association of NSW
Mr Ted Moore, Coles Supermarkets, Melbourne
Dr John Nightingale, University of New England
Dr Chris O'Donnell, University of New England
Dr Kate Owen, University of Sydney
Professor Roley Piggott, University of New England
Mr Todd Ritchie, National Farmers' Federation
Ms Kate Warden, NSW Farmers' Association
55
7. References
Ailawadi, K. L., Borin, N. and Farris, P. W. (1995). Market power and performance: a cross-industry
analysis of manufacturers and retailers. Journal of Retailing 71(Fall), 211-248.
Anon (1994). Pyrrhus at the checkout: British food retailing. The Economist 333(7884): October 8, 78.
Anon (1998). Consumer paradise. Time International 150(44): June 29, 46.
Anon (1999). Retailing: selling power. The Economist 351(8114): April 10, 75.
Ansoff, H. I. Corporate Strategy. (1965). New York: McGraw Hill.
Australian Competition and Consumer Commission (1999). Submission to the Joint Select Committee on
the Retailing Sector, Canberra.
Australian Parliament (1999). Report of the Joint Select Committee on the Retailing Sector,
(www.aph.gov.au/senate/committee/retail_ctte/retail/), Canberra.
Bain, J. S. (1951). Relation of profit rate to industry concentration: American manufacturing, 1936-40.
Quarterly Journal of Economics 65: 293-324.
Bain, J. S. (1956). Barriers to New Competition. Cambridge, Massachusetts: Harvard University Press.
Baumol, W. J. and Sidak, J. G. (1995). Toward Competition in Telephony. Cambridge, Massachusetts:
MIT Press.
Baumol, W. J., Panzar, J., and Willig, R. (1982) Contestable markets: An uprising in the theory of
industry structure. American Economic Review 72:1-15.
Baumol, W. J., Panzar, J., and Willig, R. (1982) Contestable Markets and the Theory of Industry
Structure. New York: Harcourt Brace Jovanovich.
Beeby, M. (2000). Aldi to fight big chains on price. www.afr.com.au/investment, April 3.
Brander, J. (1992). Recent Developments in Industrial Organisation. Proceedings of the Conference of
Industry Economics. Canberra: Bureau of Industry Economics. 1-25.
Bresnahan, T. F. (1989). Empirical studies of industries with market power. In: Schmalensee, R. and
Willig, R. (eds), Handbook of Industrial Organisation. Amsterdam: North-Holland.
Chamberlin, E. H. (1933). The Theory of Monopolistic Competition. Cambridge, Massachusetts: Harvard
University Press.
Chang, H-S. and Griffith, G. R. (1998). Examining long-run relationships between Australian beef prices.
Australian Journal of Agricultural and Resource Economics 42(4), 369-388.
Coase, R. H. (1937). The Nature of the Firm. Economica 4 (new series):386–405.
Collins, N. R. and Preston, L. E. (1966). Concentration and price-cost margins in food manufacturing
industries. Journal of Industrial Economics 14(3): 226-242.
Connor, J. M., Rogers, R. T., Marion, B. W. and Mueller, W. F. (1985). The Manufacturing Industries:
Structure, Strategies, Performance, and Policies. Lexington, Kentucky: Lexington Books.
56
Corbett, P. (1998). Recent changes in price relationships in the beef market, B.Ag.Ec. dissertation,
University of New England, Armidale.
Cotterill, R. (1986). Market power in the retail food industry: evidence from Vermont. Review of
Economics and Statistics. 68: 379-386.
Cotterill, R. W. (1997). The food distribution system of the future: convergence towards the US or UK
model. Agribusiness 13(2): 123-135.
David, P. A. (1985). Clio and the Economics of QWERTY. American Economic Review 75 (2): 332-37.
Dawkins, P, Bosworth, D, Harris, Michael, Harris, Mark, McDonald, E., and Rogers, M. (1998).
Innovation, Productivity And Profitability of Australian Enterprises. Melbourne: Institute of
Applied Economic and Social Research, University of Melbourne.
Dawson, P. J. and Tiffin, R. (1997). Estimating marketing margins in the meat sector using cointegration
analysis. Agricultural Economics Society Annual Conference, University of Edinburgh, 21-24
March.
Downie, J. (1958). The Competitive Process. London: Duckworth.
Duke, R. (1992). Structural changes in the UK grocery retail market. British Food Journal 94(2): 18-23.
Evans, D. S. (1983). Breaking up Bell. New York: North-Holland.
Financial Times (1999). May 12.
Fisher, B. S. (1981). A change in marketing charges: who loses what? American Journal of Agricultural
Economics. 63(2): 261-263.
Foss, N. J. (1998). Post-Marshallians and Austrians: the bridging work of G. B. Richardson. In: Foss, N. J.
and Loasby, B. J. (eds), Capabilities and Coordination: Essays in Honour of G B Richardson.
London: Routledge, 138-162.
Galbraith, J. K. (1952). American Capitalism: The Concept of Countervailing Power. London: Hamish
Hamilton.
Gardner, B. (1975). The farm-retail price spread in a competitive food industry. American Journal of
Agricultural Economics 57(3): 399-409.
Goodwin, B. K. and Holt, M. T. (1999). Price transmission and asymmetric adjustment in the US beef
sector. American Journal of Agricultural Economics 81(3): 630-637.
Griffith, G. R. and Gill, R. A. (1984). The Effects of Concentration and Vertical Integration on Profits and
Margins: A Review. Miscellaneous Bulletin No. 45. Sydney: Department of Agriculture New
South Wales, Division of Marketing and Economic Services.
Griffith, G. R., Gill, R. A. and Burgess, D. M. N. (1974). Pig industry concentration: impact on pigmeat
price spreads. In: Proceedings of the LGPA National Workshop “Marketing of Pigs and
Pigmeat”. Sydney: Livestock and Grain Producers Association.
Hall, L., Schmitz, A. and Cothern, J. (1979). Beef wholesale-retail marketing margins and concentration.
Economica 46(3): 295-300.
Harberger, A. (1954). Monopoly and Resource Allocation. American Economic Review 44 (May): 77-87.
57
Hayek, F. (1945). The Use of Knowledge in Society. American Economic Review 35: 519-30.
Hazeldene, T. (1978). Structure, Cycles and Economic Performance in the Canadian Food Manufacturing
Industry. Working Paper. Ottawa: Economic Council of Canada.
Helmberger, P. G. and Hoos, S. (1965). Co-operative Bargaining in Agriculture. Berkeley, California:
University of California, Division of Agricultural Sciences.
Helmberger, P. G., Campbell, G. R. and Dobson, W. D. (1981). Organisation and performance of
agricultural markets. In: Martin L. R. (ed), A Survey of Agricultural Economics Literature,
Volume 3. Minneapolis: University of Minnesota Press (for the AAEA): 503-653.
Herr, W. McD. (1964). The changing structure of the food processing industry. Economic Record. 40(91):
412-425.
Holloway, G. J. (1991). The farm-retail price spread in an imperfectly competitive food industry.
American Journal of Agricultural Economics. 73(4): 979-989.
Hyde, C. E. and Perloff, J. M. (1997). Multimarket market power estimation: the Australian retail meat
sector. 41st Annual Conference of the Australian Agricultural and Resource Economics Society.
Gold Coast, January.
Imel, B. and Helmberger, P. G. (1971). Estimation of structure-profits relationships with application to the
food processing sector. American Economic Review. 61(4): 614-627.
Iskow, J. and Sexton, R. (1992). Bargaining Associations in Grower-Processor Markets for Fruits and
Vegetables. ACS Research Report Number 104. Washington, DC: Agricultural Cooperative
Service, US Department of Agriculture.
King, S. and Maddock, R. (1996). Unlocking the Infrastructure: The Reform of Public Utilities in
Australia. Sydney: Allen & Unwin.
Kinsey, J. D. (1998). Concentration of ownership in food retailing: a review of the evidence about
consumer impact. Working paper 98-04. St Paul, Minnesota: The Retail Food Industry Center,
University of Minnesota.
Kinsey, J. D., Senaur, B., King, R. P. and Phumpiu, P. F. (1996). Changes in retail food delivery: signals
for producers, processors and distributors. Working paper 96-03. St Paul, Minnesota: The Retail
Food Industry Center, University of Minnesota.
Kirzner, I. (1973). Competition and Entrepreneurship. Chicago: University of Chicago Press.
Koontz, S. R., Garcia, P. and Hudson, M. A. (1993). Meatpacker conduct in fed cattle pricing: an
investigation of oligopsony power. American Journal of Agricultural Economics 75(3): 537-548.
Lamm, R. McF. (1981). Prices and concentration in the food retailing industry. Journal of Industrial
Economics. 30(1): 67-78.
Loasby, B. J. (1976). Choice, Complexity and Ignorance. Cambridge, UK: Cambridge University Press.
Loasby, B. J. (1989a). Joan Robinson’s ‘wrong turning'. In: Loasby, B. J. The Mind and Method of the
Economist. Cheltenham, UK: Edward Elgar: 71-85.
Loasby, B. J. (1989b). Knowledge and organisation: Marshall’s theory of economic progress and
coordination. In: Loasby, B. J. The Mind and Method of the Economist. Cheltenham, UK:
Edward Elgar: 47-70.
58
Loasby, B. J. (1991). Equilibrium and Evolution. Manchester: Manchester University Press.
London Economics (1997). Competition in Retailing, Research Paper 13. Prepared for the Office of Fair
Trading. London: Office of Fair Trading, September.
Lopez, R. E. (1984). Measuring oligopoly power and production responses of the Canadian food
processing industry. Journal of Agricultural Economics 35: 219-230.
Manchester, A. C. (1974). Market Structure, Institutions and Performance in the Fluid Milk Industry.
Agricultural Economics Report 248. Washington, DC: US Department of Agriculture.
Marion, B. W. et al. (1979). The price and profit performance of leading food chains. American Journal of
Agricultural Economics 59(3): 420-433.
Mason, E. S. (1939). Price and Production Policies of Large-scale Enterprise. American Economic Review
29: 61-74.
Mason, E. S. (1949). The current state of the Monopoly Problem in the United States. Harvard Law
Review 62: 1265-1285.
Messinger, P. R. and Narasimhan, C. (1995). Has power shifted in the grocery channel? Marketing
Science 14(2): 189-223.
Miller, S. E. (1982). The structural stability of the concentration-performance relationship in food
manufacturing. Southern Journal of Agricultural Economics 14(2): 43-49.
Miller, S. E. and Harris, H. M. (1983). Monopsony Power in Livestock Procurement: The Case of
Slaughter Hogs. Working Paper WP121483. Clemson, South Carolina: Department of
Agricultural Economics and Rural Sociology, Clemson University.
Mitchell, S. (2000a). German giant has time, money to win Australia. www.afr.com.au/investment, April
3.
Mitchell, S. (2000b). Weston fined $900,000 for price fixing. www.afr.com.au/investment, May 31.
Muth, R. F. (1965). The derived demand for a productive factor and the industry supply curve. Oxford
Economic Papers 16: 222-234.
Muth, M. K. and Wohlgenant, M. K. (1999a). Measuring the degree of oligopsony power in the beef
packing industry in the absence of marketing input quantity data. Journal of Agricultural and
Resource Economics 24(2): 299-312.
Muth, M.K. and Wohlgenant, M. K. (1999b). A test for market power using marginal input and output
prices with application to the US beef processing industry. American Journal of Agricultural
Economics 81(3): 638-643.
National Commission on Food Marketing (1996), Food From Farmer to Consumer. Washington DC: US
Government Printing Office.
National Farmers Federation (1999). Submission to the Joint Select Committee on the Retailing Sector.
Canberra: National Farmers' Federation.
Nelson, R. R. and Winter, S. (1982). An Evolutionary Theory of Economic Change. Cambridge,
Massachusetts: Belknap Press.
59
Nightingale, J. J. (1996). Evolutionary Processes and Revolutionary Change. In: Management, Marketing
and the Competitive Process. Earl, P. E. (ed) Cheltenham, UK: Edward Elgar: 8-46.
Norman, N. R. (1971). Manufacturing Industry in Australia. Canberra: Australian Industries
Development Association.
O’Donnell, C. J. (1999). Marketing margins and market power in the Australian dairy processing and
retailing sectors. 43rd Annual Conference of the Australian Agricultural and Resource
Economics Society. Christchurch, January.
Office of Fair Trading (1999). Bridgeman Refers Supermarkets. Press release PN 11/99, April 8.
Palaskas, T. B. (1995). Statistical analysis of price transmission in the European Union. Journal of
Agricultural Economics 46(1): 61-69.
Park, T. and Weliwite, A. (1999). Competitive behaviour in the US food retailing industry. Canadian
Journal of Agricultural Economics 47(1): 45-56.
Paul, C. J. M. (1999a). Market power measures and their uses: where's the costs? Industry Economics
Conference. Melbourne, August.
Paul, C. J. M. (1999b). Aggregation and the measurement of technological and market structure: the case
of the US meat packing industry. American Journal of Agricultural Economics 81(3): 624-629.
Penrose, E. T. (1959). Theory of the Growth of the Firm. Oxford: Basil Blackwell.
Phlips, L. (ed) (1998). Applied Industrial Economics. Cambridge, UK: Cambridge University Press.
Piggott, R. R. (1970). Elasticity of supply and the incentive for collusive buying. Australian Journal of
Agricultural Economics 14(2): 131-7.
Porter, M. E. (1980). Competitive Strategy: Techniques for Analysing Industries and Competitors. New
York: The Free Press.
Porter, M. E. (1985). Competitive Advantage: Creating and Sustaining Superior Performance. New York:
The Free Press.
Porter, M. E. (1990). The Competitive Advantage of Nations. London: Macmillan.
Prices Surveillance Authority (1994). Drought! The impact on retail food prices, Price Probe, Issue No.
20, September-December, 12-14.
Productivity Commission (1999). Impact of Competition Policy Reforms on Rural and Regional
Australia. Canberra: Commonwealth of Australia.
Putterman, L. and Krosner, R. (1996). The Economic Nature of the Firm, 2nd ed. Cambridge, UK:
Cambridge University Press.
Queensland Fruit and Vegetable Growers (1999). Submission to the Joint Select Committee on the
Retailing Sector. Brisbane: Queensland Fruit and Vegetable Growers Association.
Richardson, G. B. (1990). The Organisation of Industry. In: Richardson, G. B. Information and
Investment, 2nd ed. Oxford: Oxford University Press: 225-242.
Rizvi, S. M. H. and Uhm, I. H. (1982). Market structure and economic performance of the food
processing and farm input industries in Canada. Canadian Farm Economics 17(4): 1-8.
60
Roberts, J. (1984). Testing oligopolistic behaviour. International Journal of Industrial Organisation 2:
367-383.
Robinson, J. (1933). The Economics of Imperfect Competition. London: Macmillan.
Round, D. K. (1974). The industrial organisation vacuum in Australia. Economic Record 50(100): 169198.
Scherer, F. M. (1980). Industrial Market Structure and Economic Performance, 2nd ed. Chicago: Rand
McNally.
Scherer, F. M. and Ross, D. (1990). Industrial Market Structure and Economic Performance, 3rd ed.
Boston: Houghton Mifflin.
Schroeter, J. (1988). Estimating the degree of market power in the beef packing industry. Review of
Economics and Statistics 70(1): 158-162.
Schroeter, J. and Azzam, A. (1991). Marketing margins, market power and price uncertainty. American
Journal of Agricultural Economics 73(4): 990-999.
Shepherd, W. G. (1995). Contestability vs. competition - once more. Land Economics 71(3): 299-309.
Sheridan, K. (1974). The Firm in Australia. Melbourne: Nelson.
Simon, H. A. (1972). Theories of Bounded Rationality. In: Radner, C. B. and Radner, R. (eds), Decision
and Organisation. Amsterdam: North Holland: 405-23.
Slade, M. E. (1995). Beer and the tie: did divestiture lead to higher prices? Discussion Paper 95/44.
Vancouver: Department of Economics, University of British Columbia, November.
Smith, R. and Round, D. (1998). Competition Assessment and Strategic Behaviour. European
Competition Law Review 19(4): 225-236.
Sosnick, S. (1958). A Critique of Concepts of Workable Competition. The Quarterly Journal of
Economics 72 (3): 380-423.
Sutton, J. (1991). Sunk Costs and Market Structure. Cambridge, Massachusetts: MIT Press.
Tomek, W. G. and Robinson, K. L. (1990). Agricultural Product Prices, 3rd ed. Ithaca, New York: Cornell
University Press.
US Department of Agriculture (1996). Concentration in Agriculture. Report of the Advisory Committee
on Agricultural Concentration. Washington, DC: Agricultural Marketing Service, US Department
of Agriculture.
US Department of Agriculture (1999a). Concentration and Competition in the U.S. Food and Agricultural
Industries. Agricultural Outlook. Washington, DC: Economic Research Service, US Department
of Agriculture, May, 26-29.
US Department of Agriculture (1999b). More farmers contracting to manage risk. Agricultural Outlook.
Washington, DC: Economic Research Service, US Department of Agriculture, January/February,
6-7.
US Federal Trade Commission (1969). Economic Report on the Influence of Market Structure on the
Profit Performance of Food Manufacturing Industries. Washington, DC: US Government
Printing Office.
61
Vernon, J. M. (1972). Market Structure and Industrial Performance: A Review of Statistical Findings.
Boston: Allyn and Bacon.
Wann, J. J. and Sexton, R. J. (1992). Imperfect competition in multiproduct food industries with
application to pear processing. American Journal of Agricultural Economics 74(4): 980-990.
Weiss, L. (1971). Quantitative studies in industrial organisation. In: Intrilligator, M. (ed), Frontiers of
Quantitative Economics. North Holland: Amsterdam: 363-403.
Weiss, L. (1974). The concentration-profits relationship and antitrust. In: Goldman, H. J. et al (eds),
Industrial Concentration: The New Learning. Toronto: Little, Brown and Co., for Columbia
University Centre for Law and Economic Studies, 184-232.
Weiss, L. (1990). Concentration and Prices. Cambridge, Massachusetts: MIT Press.
Williamson, O. E. (1985). The Economic Institutions of Capitalism. New York: Free Press.
Williamson, O. E. and Winter, S. (eds) (1991). The Nature of the Firm: Origins, Evolution and
Development. Oxford, UK: Oxford University Press.
Wohlgenant, M. K. (1999). Marketing margins: empirical analysis. Draft chapter for the American
Association of Agricultural Economics Handbook. Raleigh, North Carolina: Department of
Agricultural and Resource Economics, North Carolina State University.
Wohlgenant, M.K. and Haidaicher, R.C. (1989) Retail to Farm Linkage for a Complete Demand System
of Food Commodities. Technical Bulletin No. 1775, Washington, DC: Economic Research
Service, US Department of Agriculture.
Wrigley, N. (1997). Exporting the British model of food retailing to the US: implications for the EU-US
food system convergence debate", Agribusiness 13(2): 137-152.
Zhao, X., Griffith, G. R. and Mullen, J. D. (1997). The competitive structure of the Australian
beef industry: accounting for trade. Armidale, NSW: Unpublished paper, NSW Agriculture.
62