Two Studies of the Australian Wheat Board: a Traditional Price Discrimination Model, and the Privatisation Process and Pricing Behaviour of a Risk Averse Firm

Two studies of the Australian Wheat Board: A traditional price discrimination model, and the privatisation process and pricing behaviour of a risk averse firm.

Alexandra E. Lobb

This thesis is presented for the degree of Doctor of Philosophy of The University of Western Australia, School of Agricultural and Resource Economics, 2003

ABSTRACT

This thesis is motivated by the impacts of contemporary political and economic issues such as microeconomic reform and regulatory control on the Australian wheat industry. Firstly, the suggestion of whether the AWB (International) Ltd commands market power and secondly, that the objectives of the AWB Ltd have changed since semi-privatisation of the Australian Wheat Board under the Wheat Marketing Act, 1989.

The AWB (International) Ltd’s ability to price discriminate is a key component to the retention of the single desk regulatory arrangement for the export of Australian wheat. Due to data restrictions the market power of the AWB (International) Ltd has not been determined within this thesis.

To complement this traditional approach, a more novel proposal is developed to determine the effect of microeconomic reform on the Australian wheat industry. Conceptualising the change of the AWB Ltd’s objectives as a shift from revenue maximization to profit maximization, this study examines the impact of such a change on the pricing policies of a multi-market price-setting firm. More specifically, this study investigates, for two hypothetical objective functions, a risk averse firm’s price-setting behaviour in an “overseas” and a “domestic” market, given differing costs of supply, uncertain demand functions and differing price elasticities of demand in each market. The aim is to generate empirically testable hypotheses relating to the impact of a change of objectives on pricing behaviour.

TABLE OF CONTENTS

Abstract...... ii

Table of Contents...... iii

Table of Figures and Tables ...... vii

Acknowledgements...... ix

Chapter1 Introduction...... 1

1.1 Introduction...... 1

1.2 Objectives...... 2

1.3 Methodology ...... 3

1.4 Significance of Results and Implications...... 3

1.5 Structure of the Thesis...... 3

Chapter 2 Literature and Policy Review of the Australian Wheat Industry ...... 5

2.1 Introduction...... 5

2.2 The Political Economy of the International Wheat Market...... 5 2.2.1 The Demand for Grain – A General Overview...... 6 2.2.2 State Trading Enterprises...... 11 2.2.3 Modelling the World Wheat Market ...... 21 2.2.4 The Market Power Debate...... 25 2.2.5 Conclusion...... 29

2.3 The Political Economy of the Australian Wheat Industry ...... 29 2.3.1 Historical background ...... 29 2.3.2 The Wheat Industry Stabilisation Act (1948) ...... 30 2.3.3 Developments in the late 1970s and early 1980s...... 31 2.3.4 The Wheat Marketing Act (1984)...... 32 2.3.5 Amendments to the Wheat Marketing Act (1989, 1992)...... 33 2.3.6 The Transition Period (1990-1997)...... 34

iii

2.3.7 The Formation of AWB Ltd, and it’s Corporate Structure ...... 37 2.3.8 The National Competition Policy Review 2000 ...... 38

2.4 The Future of the Australian Wheat Industry ...... 44

2.5 Conclusion...... 44

Chapter 3 A Traditional Analysis of a Price Discriminating Monopolist...... 46

3.1 Introduction...... 46

3.2 ACG Methodology...... 47

3.3 The Carter – Knetter Price Discrimination Model...... 49 3.3.1 Introduction...... 49 3.3.2 Review of Carter et al (1999) ...... 49 3.3.3 Results ...... 51

3.4 Functional Form of the Demand Curve ...... 55 3.4.1 The Linear Demand Model for Price Discrimination...... 56 3.4.2 Uncertain Demand Functions and the Assumption of Perfect Information..58 3.4.3 Data ...... 59 3.4.4 Results ...... 60 3.4.5 Conclusion...... 63

3.5 Systematic Price Premiums ...... 64 3.5.1 The Application of the Carter-Knetter Model...... 64 3.5.2 The Revised Equilibrium Model...... 67 3.5.3 Data ...... 69 3.5.4 Results ...... 71 3.5.5 Conclusion...... 75

3.6 Conclusion...... 76

Chapter 4 A Political Economic Review of Microeconomic Reform and Privatisation 78

4.1 Introduction...... 78

4.2 Microeconomic Reform and The Privatisation Process...... 78

4.3 Conclusion...... 84

Chapter 5 Modelling the Behaviour of the Australian Wheat Board...... 85

5.1 Introduction...... 85

5.2 Assumptions of the Model...... 86 5.2.1 Wheat as a homogenous good ...... 86 5.2.2 Three markets ...... 87 5.2.3 Costs...... 87 5.2.4 Elasticity of demand ...... 88 5.2.5 Uncertainty of demand...... 89 5.2.6 AWB Ltd as a price setting firm...... 89 5.2.7 AWB Ltd as a risk averse firm...... 90

5.3 The Model...... 91

5.4 Numerical Analysis ...... 96 5.4.1 Sensitivity Analysis...... 100 5.4.2 Demand Functions...... 101 5.4.3 Hypotheses...... 103

5.5 Conclusion...... 103

Chapter 6 Implications of Recent Industry Developments for Domestic and Overseas Prices 105

6.1 Introduction...... 105

6.2 Domestic Deregulation of the Australian Wheat Market...... 105 6.2.1 Results ...... 108

6.3 Changes in Transport Costs...... 110 6.3.1 Domestic Costs...... 111 6.3.2 Export Costs...... 115 6.3.3 Data – domestic costs...... 118

6.3.4 Results ...... 122

6.4 Uncertainty in the International Arena ...... 124 6.4.1 Results ...... 128

6.5 Sensitivity Analysis...... 131

6.6 Conclusion...... 133

Chapter 7 Empirical Analysis...... 135

7.1 Introduction...... 135

7.2 Data Set...... 135

7.3 Price Relationships ...... 138 7.3.1 Application to the Generated Hypotheses...... 142

7.4 Extension to the model...... 145 7.4.1 Results ...... 148

7.5 Conclusion...... 150

Chapter 8 Conclusion...... 151

8.1 Introduction and Summary...... 151

8.2 Key Findings and Contributions ...... 152

8.3 Limitations and Further Research ...... 155

Appendix 1 Australia’s Key Import Markets...... 156

Appendix 2 Regression Analysis Results...... 163

Appendix 3 Australian Wheat Quality Characteristics...... 177

References...... 187

TABLE OF FIGURES AND TABLES

Table 2.1 The Big Four - wheat exporter’s market share...... 8 Table 2.2 Major importers of Australian wheat...... 11 Figure 2.1 World wheat exports...... 13 Table 2.3 Producer Support Estimates...... 20 Table 2.4 Evidence of the Market Power of the AWB...... 27 Figure 2.2 The ‘Grower Corporate Model’...... 37 Table 3.1 Simulated average price premia derived across all countries...... 61 Table 3.2 Simulated price premia ...... 62 Table 3.3 Imputed elasticities of demand prices ...... 63 Table 3.4 Premiums received by class of wheat...... 71 Table 3.5 Total value of premium by class...... 72 Table 3.6 Premiums received by class ...... 73 Table 3.7 Total value of premium ...... 74 Figure 3.1 Price premiums received by class ...... 75 Table 5.1 Simulated results for prices and quantities as a result of a change in objectives 97 Table 5.2 Simulated results for prices and quantities as a result of a change in objectives 98 Table 5.3 Simulated results for prices and quantities as a result of a change in objectives 99 Table 5.4 Sensitivity to changes in the relative risk aversion coefficient...... 100

Table 6.1 Comparing sales maximisation results when bd is increased ...... 108

Table 6.2 Comparing profit maximisation results when bd is increased ...... 109 Figure 6.1 Domestic freight transport ...... 111 Figure 6.2 International port authority charges for bulk wheat exports...... 117 Table 6.3 Real rail freight price trends...... 118 Table 6.4 Port authority costs...... 120 Table 6.5 Comparing sales maximisation results when costs are decreased ...... 122

vii

Table 6.6 Comparing profit maximisation results when costs are decreased ...... 123 Figure 6.3 Average annual gold prices...... 127 Figure 6.4 Price index for world gold and wheat prices ...... 128 Table 6.7 Comparing sales maximisation results when uncertainty in the overseas market decreases...... 129 Table 6.8 Comparing profit maximisation results when uncertainty in the overseas market decrease...... 130

Table 6.9 Comparing revenue maximisation results when bd is increased and costs have declined in both markets...... 132

Table 6.10 Comparing profit maximisation results when bd is increased and costs have declined in both markets...... 132 Table 6.11 Comparing revenue maximisation results with the expected profit constraint set at 90% of maximum expected profits...... 133 Table 7.1 Australian wheat data...... 137 Table 7.2 Wheat prices...... 138 Figure 7.1 World price and Australian export price for wheat...... 139 Table 7.3 Wheat prices and the price ratio ...... 140 Figure 7.2 World wheat price and Australian domestic wheat price ...... 141 Table 7.4 Price, quantity and expected profit, “before” and “after” - values using real data set 143 Figure 7.3 World wheat use...... 145 Figure 7.4 Australian export and domestic wheat prices and the price ratio ...... 147 Table 7.5 Price, quantity and expected profit, “before” and “after” - the base case scenario 149 Table 7.6 Price, quantity and expected profit, “before” and “after” - a decline in world price for the Australian Wheat Board...... 149

viii

ACKNOWLEDGEMENTS

Firstly, I would like to thank the Grains Research and Development Corporation for funding this project, the AWB Ltd for permitting the use of their data set, and the University of Western Australia who allowed me to complete my thesis as an external student. Secondly, for the support of my supervisors, Professor Rob Fraser, Imperial College, Wye, a most respected mentor who provided much support, and Associate Professor Michael Burton, University of Western Australia, for their invaluable time and expertise in all areas of my academic life. Finally, I would like to thank my husband, Alistair McEwan, and my parents, Wendy and Hilton Lobb, for their constant encouragement.

CHAPTER1 INTRODUCTION

1.1 INTRODUCTION

Wheat is seen as one of the world’s leading food grains. Wheat is consumed and produced in nearly every country worldwide. International wheat trade has increased from 41.9 million tonnes in 1960 to 100 million in 1998, a 250% increase. The significance in agricultural commodity markets partly accounts for the sometimes intense political pressures and interventions to which wheat is subject. Australia produced an average of 22 million tonnes of wheat a year over the last five crops years (1996/97 to 2000/01), with an average of 76% of this exported each year. The current value (2001) of Australia’s wheat exports are over A$3 billion, representing 4% of total Australian export goods and services. Australia is the third largest wheat exporting nation, behind the USA and Canada, closely followed by the EU and Argentina.

Australia’s wheat exports have been marketed under the Australian Wheat Board (AWB), the sole export agent, since 1939. As a result, the Australian wheat industry has historically been the beneficiary of considerable government-funded support. However, commencing with the cessation of the Guaranteed Minimum Price Scheme in 1989 this support has been in the process of being removed, with the aim of leaving the industry exposed to economic realities. Over this period the central player in the Industry has been the former Australian Wheat Board (AWB), and its activities have been particularly targeted in relation to the removal of government-funded support and the encouragement to adopt fully commercial practices. However, through AWB International Ltd. (AWB(I)), there remains a single desk arrangement for the marketing of Australian wheat for export.

The single desk regulation implies that the AWB(I) acts as a state trading enterprise (STE) on the export market. As a result, it is plausible that the marketing activities of the AWB(I) are having an adverse effect on international trade. STEs are believed to have distortionary

1 effects on trade and are currently on the World Trade Organisation’s (WTO) agenda for future trade rounds and negotiations. The central issue is to determine if AWB(I) has the ability to price discriminate because of its statuatory control over Australian wheat exports, and hence holds market power which may mean that other wheat trading nations are not able to compete with Australian wheat in the global arena.

Given an ability to price discriminate, the structural policy changes imposed on the AWB as a result of microeconomic reform have been similar to those imposed by governments on other former monopoly public enterprises in the privatisation process. This process has been the subject of a considerable economics literature, with one of the focuses of this literature being the impact of privatisation on the objectives of the firm, and consequently on its behaviour (Fraser, 1989; Vickers and Yarrow, 1989; Bishop et al., 1994; Fraser, 1991; Fraser, 1994(a) and Fraser, 1996). However, one key difference is that whereas the privatisation of public enterprises that retain considerable monopoly power have been associated with post-privatisation regulation of their behaviour, typically of the “price-cap” variety, the AWB Ltd, by virtue of trading across national boundaries, is not subjected to any price regulation.

This observation raises the question of whether an examination of the AWB Ltd’s situation using the methods of the privatisation literature might reveal insights regarding how its behaviour is likely to have been modified by the removal of government-funded support to the Industry.

1.2 OBJECTIVES The aims of this thesis are two fold. Firstly, to examine whether the existence of the continued support by the government for the AWB Ltd’s single desk operations are founded as a price setter, and secondly, to undertake an analysis of the AWB’s behaviour, focusing in

2 particular on how the government’s push of the AWB toward fully commercial practices can be expected to have affected its pricing behaviour, given the AWB Ltd’s ability to set prices.

1.3 METHODOLOGY Due to the bilateral nature of this investigation two models are developed. First, a price discrimination model based on the Carter-Knetter framework is developed and applied empirically using regression analysis to determine if the AWB Ltd’s single desk operations may command price premia for Australian wheat. Second, given this a theoretical model is developed to examine the pricing behaviour of a firm under different objectives in order to represent the Australian government’s push of the AWB Ltd towards fully commercial practices. This model is adapted from that of Fraser (1989) and numerical results and extensions are presented.

1.4 SIGNIFICANCE OF RESULTS AND IMPLICATIONS This thesis expands the academic research on the Australian wheat industry. Of particular significance is the presentation of a novel theoretical analysis of the AWB Ltd’s pricing behaviour as its objectives change as a result of developments following microeconomic reform. Such an analysis had not been undertaken prior to this study nor has the application of a multi-market, risk averse firm with market power been made to the modelling processes in the privatisation literature.

1.5 STRUCTURE OF THE THESIS The structure of this thesis is as follows. Chapter 2 provides a review of the literature concerning the policy of the international and Australian wheat industry focusing on the demand for wheat, policy implications of institutions such as state trading enterprises and the market power debate and an in depth analysis of the political economy of the Australian wheat industry. Chapter 3 follows from the final section in chapter 2 and develops a traditional price discrimination modelling approach based on research conducted as part of the 2000 National Competition Policy Review of the Wheat Marketing Act. Chapter 4

3 presents a literature and policy review of the privatisation process, with a focus on microeconomic reform in Australia. Chapter 5 presents an extension of a theoretical model based on that of Fraser (1989) of a size-orientated price-setting firm operating in multiple markets. The assumptions of the model are outlined and validated, the model is presented algebraically, and an initial numerical analysis is reported. Chapter 6 further expands the model by imposing recent Australian wheat industry developments including deregulation of the domestic Australian wheat market; changes in domestic and export transport costs; and developments in international stability. Chapter 7 presents an empirical study, somewhat constrained due to the lack of available data, and chapter 8 suggests areas of further research and concludes the thesis.

CHAPTER 2 LITERATURE AND POLICY REVIEW OF THE AUSTRALIAN WHEAT INDUSTRY

2.1 INTRODUCTION Chapter 2 provides a political economic background to the international and the Australian wheat markets. Section 2.2 outlines the structure of the world wheat market and addresses issues such as state trading enterprises and market power within the industry. Section 2.3 provides an historical background to the Australian wheat market with specific focus on the former Australian Wheat Board, the domestic market deregulation, and the privatisation process and the recent National Competition Policy (NCP) Review. Section 2.4 investigates future issues for the Australian wheat industry and section 2.5 concludes.

2.2 THE POLITICAL ECONOMY OF THE INTERNATIONAL WHEAT MARKET One of the most important components of export markets for agricultural goods is government involvement in production and domestic or international marketing processes which may have a flow through effect on trade. This involvement may include marketing arrangements such as single desk exporting, subsid ized trade and credit programmes. International aid controlled by governments may also impact on demand for wheat around the world.

The former Australian Wheat Board’s single desk, AWB(I), claimed command of market power has been central to the debate surrounding the 2000 NCP Review of the Wheat Marketing Act (1989), (WMA) with reference to the single desk and, more widely, is the key to World Trade Organisation (WTO) discussions on the powers of State Trading Enterprises (STEs).

2.2.1 THE DEMAND FOR GRAIN – A GENERAL OVERVIEW Throughout the late 1990s the world’s grain markets were depressed, with relatively weak demand and low prices. Forecasters predict an increase in grain consumption over the next few years which should offset the continued increases in global production (Turner et al., 2000, p 31). The main drivers for this rise in demand include an increase in world economic growth and the resulting changing consumption patterns, specifically in Asia, the Middle East and South America. The grains industry is set to benefit both directly and indirectly as a result of rising incomes per capita and associated effects. Firstly, through increased market potential for wheat and related end use products, such as bread and noodles, and secondly through the export of livestock feed to fuel the rising demand for meat and poultry in these regions. See Antle and Smith (1999) for an overview of the international wheat market.

Wheat is often perceived as a heterogeneous good, as quality characteristics and suitability for end use products is a key component of demand for wheat. Although there is trading competition for wheat in general, due to the many different varieties of wheat grown around the world there is much more competition within specific grades and between substitutable grades. For example noodle production requires white wheat with good milling properties, low moisture content, required level of starch damage and superior dough strength (e.g. Australian Prime Hard or Australian Premium White). The USA competes with Australia in the production of wheats suitable for noodles although the US predominantly produces red wheats and the substitution between these and the Australian varieties is low. However, with the advent of new milling, blending, production technology and biotechnology the differences between qualities are less enhanced than in previous decades.

Turner et al., (2000) highlight in “Grains – Outlook to 2004-05”, that there are three forces over the medium term which may have a significant impact on international grain trade: Policy developments in major grain producing and exporting countries; Potential for an increase in production for developed and developing countries through the introduction of

6 new technologies; and Potential export market expansion in Asia, the Middle East and North Africa due to changing consumption patterns (Turner et al., 2000, p 35).

In 1993, at the Uruguay Round, the World Trade Organisation (WTO) was formed (following on from the General Agreement on Tariffs and Trade (GATT)). This was the first time that agricultural trade had been discussed in the world arena. Although the gains were nominal, the Uruguay Round set the stage for further developments in creating freer trade for agricultural goods (ERS, 1998). As a part of these developments it was suggested that some of the trade policies of large exporters such as the USA and the EU were hindering free trade (Roberts and Doyle, 1996). These adverse effects were mainly due to price distorting and market sheltering programmes that gave the EU and the US an unfair advantage. It was recommended that all countries, both developed and developing, alter their trade practices over the period 1993 – 2000, in accordance with certain guidelines announced by the WTO in which policies should be changed to meet specific requirements or be discarded (WTO, 1994, 1995). All guidelines ensured that there would be a satisfactory adjustment period for the policy recipient. It is important to note that this has only been accomplished in a limited capacity given antagonisms between nations and domestic protests in member countries. A key component of future WTO rounds is the presence of STEs, and whether they inhibit trade (Dixit and Josling, 1997; McCorriston and MacLaren, 2002).

EXPORTERS The major exporters of wheat are the US, Canada, Australia, the EU and Argentina. In 1997/98 the US exported over 28 million tonnes of wheat, the same as Australian and EU exports combined (ABARE, 1999). Australia exports an average of 77% of all wheat produced, valued at over A$3 billion in 1997/98 (ABARE, 1999). Total Australian grain exports represent approximately 20% of the value of Australian export goods and services

(AWB, 1999). Australia exports predominantly to Asia and the Middle East, and their major competitors are the USA and Canada.

Table 2.1 The Big Four - wheat exporter’s market share

(Average 1993/94 – 2002/03)

Country % Share USA 27.7% Canada 16.3% EU 15% Australia 13.7% TOTAL 72.7%

(Source: CWB, 2003, Table 30, p 25)

Of these major exporting nations, the US and the EU have federally funded agricultural policies that are thought to have an impact on the international trade of wheat (Turner et al., 2000; Tarchalski et al., 1996; Roberts and Doyle, 1996; Anania et al., 1992). These complex programmes effectively shelter wheat producers from market conditions through direct export subsidies and market development, access and assistance (or loan) programmes.

THE CANADIAN WHEAT BOARD Canada is the world’s second largest wheat exporter behind the US, contributing to approximately 16% of world wheat trade (1993/4-2002/3) (See Table 2.1). Canada, not unlike Australia, is considered to be a supplier of ‘high quality wheat’ (approximately two thirds of wheat exports fall into the high quality category, No. 1 and 2 CWRS), although the market for high quality wheat is small and demands high specifications for protein, hardness, moisture and colour. The lack of growth in the high quality market (UK and Japan) has lead

8 to much discussion on Canada’s opportunities to increase revenue in the lower quality wheat markets such as China, Brazil, South East Asia and the Middle East (Carter, Loyns and Ahamadi-Esfahani, 1986; Ulrich, Furtan and Schmitz, 1987)

Canada’s domestic and international wheat market is controlled by the Canadian Wheat Board (CWB), a government owned and operated statutory marketing authority (state trading enterprise). The Canadian system, following from the Australian system (1915), was formed in 1919 during the first world war to maintain depressed wheat prices. The CWB was legislated by parliament in 1935, however, the CWB was subject to periodic amendments until 1967 when the parliament amended its position without expiration, thereby creating a permanent fixture (Wilson, 1989).

The CWB is a more traditional state trading enterprise than the AWB Ltd, as the CWB controls both the national and international sales and distribution of Canadian wheat, as well as less commercial and more reliant on government underwriting. The CWB consists of single desk selling, pooling regime and a Canadian government guarantee on CWB borrowings and on initial farmer payments (www.cwb.ca, 2004). The organizational structure of the CWB also differs to that of the AWB Ltd.

IMPORTERS The world’s largest importer of wheat is the EU, importing over 25 000 Kt in 1997/98, more than the next four largest importers together (ABARE, 2000, p 223). The other major wheat importing countries in the 1997/98 season included Egypt, Brazil, Japan, Iran, Algeria, Pakistan, South Korea and Indonesia. In previous years China has also been a large importer of wheat, 12 500 Kt in 1995/96 (ABARE, 2000, p 223), however, domestic production in China is unstable and hence their import regime is highly variable (see Rozelle and Huang, 1999, for detail on wheat in China).

Importing countries may also place restrictions on goods entering their country hence impacting on international trade flows. These policies may include tariffs, quotas, quality restrictions and may be motivated by either political control, revenue raising or genuine quarantine concerns.

Asia is Australia’s largest export market, importing on average over 6700 Kt of wheat per year, followed by the Middle East at 3036 Kt., and Africa at 1298 Kt per year (ABARE, 1999). Table 2.2 presents countries which import a high proportion of Australian wheat.

Table 2.2 Major importers of Australian wheat

(1991-1999) Country Average Quantity Imported (Kt) Indonesia 1546 Japan 1156 Egypt 1119 South Korea 830 India 682 China 671 Malaysia 659 Iraq 613 Pakistan 568

(Source: ABARE, 1999)

See Appendix 1 for a detailed discussion on Australia’s main import markets.

2.2.2 STATE TRADING ENTERPRISES Throughout the world there are approximately 150 bodies that are classified as agricultural export or import state trading enterprises (STEs) by the WTO, (WTO, 1995). Agricultural products account for 70% of STEs for the trade of commodities such as wheat, feed grains, sugar, rice and dairy products (McCorriston and MacLaren, 2002, p 136). Of these, the majority are in developing countries such as Tunisia, Mauritius, Jamaica or former Eastern bloc countries like Poland or the Czech Republic. In the developed world, Australia, New Zealand and Canada have significantly active export STEs and, as with import STEs which include agencies in Japan and Switzerland (Ackerman and Dixit, 1999, p 35-37).

Nearly half of the world’s wheat export and import markets are dominated by STEs. Australia and Canada both export their wheat via a Wheat Board, a statutory marketing authority (SMA), with aims to maximise producer returns. The AWB(I) and the CWB account for approximately one third of all wheat traded on the international export market (See figure 2.1, below). Many former centrally planned Eastern European nations also control exports, including Poland and Kazakhstan. Between 1994 and 1997 these Eastern European countries held an average of 6% of world wheat exports (Ackerman and Dixit, 1999, p 4). Wheat imports are also controlled by government importing agencies in countries such as, Japan, Indonesia, China, Egypt, Pakistan, and Tunisa. China, for example, imports over US$1.27 billion of wheat through their China National Cereals, Oil and Foodstuffs Import and Export Corporation (COFCO) (Ackerman and Dixit, 1999, p 8-9).

Figure 2.1 World wheat exports

(average for marketing years 1994-97)a

Eastern Europe Australia 6% USA 13% 31%

Canada 20% EU Others Argentina 16% 6% 8%

aHatched area represents nations with STEs (39% of all exports). (Source: Ackerman and Dixit, 1999, p 4).

There are many ambiguities in relation to the definitional requirements of an STE and what is “technically” an STE. The WTO defines an STE as: “governmental and non- governmental enterprises, including marketing boards, which have been guaranteed exclusive or special rights or privileges, including statutory or constitutional powers, in the exercise of which they influence through purchases or sales the level or direction of imports or exports”(WTO, 1994, p 509-511). A 1995 WTO report outlined the types of STEs that

13 exist in the world, including statutory marketing authorities, regulatory marketing boards, fiscal monopolies, canalizing agencies and foreign trade monopolies (WTO, 1995).

Throughout the 1970s state trading was seen solely as a government body monopolising foreign trade, or government ownership of an enterprise (Baldwin, 1970). This definition has since been expanded to examine issues surrounding government control of trade as it is the government control that influences the behaviour of the state trader and it is this which highlights the differences between state and private traders (Kostecki, 1982, p 22). Focus on government control and ownership has declined in recent years, specifically with the increase in microeconomic reform in developed nations. McCorriston and MacLaren (2002) comment that, “it is not ownership per se that matters but the extent to which an enterprise, even if it is a private organisation, has been granted exclusive or special rights by the government” (p 134). The AWB Ltd is cited as the perfect example, and the definition is further reviewed to suggest that “STEs arise not necessarily from ownership but from exclusive rights” (McCorriston and MacLaren, 2002, p 135).

Many STEs were initially established in ‘emergency’ situations (e.g. war), to achieve domestic policy objectives, like the Australian Wheat Board, or for food security issues such as the Japanese Food Agency. Currently they are used primarily to stabilise prices for either consumers or producers, or to take advantage of economies of scale or scope in transport, distribution, quality or foreign marketing (Ackerman and Dixit, 1999, p 11 and Brenner, 1987). The later is often cited as a reason for the retention of the AWB(I) Ltd’s single desk. Other benefits for the existence of STEs include: exploiting market power through price discrimination thus increasing domestic revenues; providing farmers with risk management through price pooling; negotiation of price premiums with single desk buyers, that is government to government trade deals; and development of niche markets and new buyers through intensive market development (Carter and Wilson, 1999, p 205).

The possible trade distortionary affects of STEs are topical in the literature. The premise is that nations with export trade monopolies may be in a superior negotiating position to those nations attempting to privately export a commodity (See Carter and Lyons, 1996). The effect of an exporter STE is that the revenues obtained by domestic producers will rise given a fixed output. This is detrimental to trade for two reasons, firstly, a decline in sales for the competing producer in the third country as a result of the trade distortionary practices of the price discriminating STE, and secondly, the higher prices received in the country with the STE will result in an increase in the supply and possibly a situation of over production. This could lead to further trade distortions if no adequate long run storage facilities are available and the product is dumped on the world market, further depressing prices (Alston and Gray, 2000).

Ackerman and Dixit (1999), established a qualitative classification of STEs into four different categories:

Type I does not have control over domestic or international trade, little potential for trade distortion; Type II focuses on the control of the domestic market but with an open international focus, low potential for trade distortion; Type III controls imports or exports but has a deregulated domestic sector, moderate potential for trade distortion; and Type IV controls both international and domestic trade, high trade distortion potential.

(Source: Ackerman and Dixit, 1999, p 17-18, from Dixit and Josling, 1997)

Under this classification structure, Ackerman and Dixit (1999) examine eight major STE importers and exporters (p 18 - 37). Focusing on exporter STEs an example for each classification level is seen below.

Type I - The USA’s Commodity Credit Corporation; Type II - The EU’s Common Agriculture Policy; Type III - The Australian Wheat Board; and Type IV - The Canadian Wheat Board.

(Source: Ackerman and Dixit, 1999, p 18-37)

The significance of STEs on recent WTO agenda coupled with the push for microeconomic reform has prompted the push for removal of STEs from international wheat and other agricultural goods markets, specifically for developed countries. Carter and Wilson (1999) comment that, “most of the economic impact of the STEs is domestic” (p 206). This follows from:

In either Canada or Australia, once the grain gets to the port , the exporting board sells in to a competitive market. It is therefore doubtful if the boards overcharge offshore customers as they claim. It is also questionable whether they undercut offshore prices to the degree some critics claim.

(Carter and Wilson, 1999, p 206).

Wilson et al (2000), go on to suggest that there would be no change in the international wheat market as a result of the disintegration of the AWB and CWB. However, if this is the case and STEs are having no impact on world trade or market shares, then what impacts, if any, are they having, and why are they cause for such discussion?

McCorriston and MacLaren (2002, pp 140 - 141) comment that when analysing state trading issues it is the ambiguity of the definition of STEs and how they interact in a possibly imperfect market that is the crux of the lack of solutions to questions such as the one posed above. Regardless of the apparent focus on STEs in recent WTO rounds, the WTO dispute

16 settlement process, a likely candidate for contending with issues such as the trade distorting impacts of STEs, has not in its history (exception being for a recent case against a South Korean import agency for beef), conducted such a case (McCorriston and MacLaren, 2002, p 140).

There has been focus on the lack of price transparency created in international wheat markets as a result of state trading both in relation to importing and more specifically exporting nations (Wilson et al, 2000). This transparency inhibits export competition in the marketplace as nations with private trading companies find it difficult to compete with STEs as they may lack information, negotiating powers, financial mechanisms, quality standards and controls.

Price transparency, a measure of information in the market, can be defined as “the extent that details of transactions made by a purchasing or selling agent are available to the public” (Furtan, 1995). Wilson et al (2000) note 3 important aspects of transparency:

1. Differs from price discrimination; 2. Reflects informational asymmetries about costs and bidding processes which have been exploited by multinationals to the disadvantage of STEs; and 3. The variability of over time of advantaged parties. Typically multinationals have been in a superior situation during the 1970s and 1980s, however, the introduction of the EEP made the US more transparent than its competitors.

(Wilson et al, 2000).

Wilson et al (2000) conclude, using a Bayesian Nash first price, sealed bid auction to identify optimal bidding behaviour with contract data, that grain buyers are influenced by the form of competition because of the way this effects prices in the bidding system. Their results show that firstly, reducing uncertainties gives rise to lower equilibrium bids and prices. Secondly, less transparent situations (e.g. where STEs participate), lead to increasing

17 bids and buying prices and resulting in higher payoffs to sellers. Thirdly, if the number of rival firms is increased, then STEs have less of a competitive advantage, lower equilibrium bids, and payoffs. This study supports the hypothesis that nations such as the US are disadvantaged in the bidding process as a result of less transparent players in the system, with results indicating that there may be advantages to a nation having a STE export arrangement. Further, results suggest that there may be implications on the international market if the AWB(I) Ltd’s single desk were to be dissolved as, Canada’s power may decline as the number of competitive rivals in the market may increase.

Interestingly, Schmitz et al. (2000) note that futures markets are the crux of price discovery where both private enterprises and STEs compete, and hence transparency should not be an issue.

Alston and Gray’s (2000) approach is to take two markets, and fixed quantity of wheat, which has to be disposed of across these two markets. If the STE is unable to price discriminate then the price charged for wheat will equate across the two market for a given equilibrium quantity. Through price discrimination, it is possible for the STE to increase its revenue by selling more into the price inelastic market and less into the more inelastic market. This will result in the optimal allocation so that the last unit sold in each market effectively increases total revenue for the STE. Given that the STE will have to decrease price in order to increase quantity to one of the markets, the marginal revenue from the additional unit of wheat sales in a given market will be less than the price at which it is sold (Alston and Gray, 2000).

The effects of an exporter STE is that the revenues obtained by domestic producers will rise given a fixed output. This is detrimental for two reasons, firstly, a decline in sales for the competing producer in the third country as a result of the trade distortionary practices of the price discriminating STE, and secondly, the higher prices received in the country with the STE will result in an increase in the supply of wheat and possibly a situation of over

18 production. This could lead to further trade distortions if no adequate long run storage facilities are available and the product is dumped on the world market, further depressing prices.

Theory suggests that a price increase to farmers as a result of price discriminatory behaviour by a monopoly is only plausible if the importer’s elasticity of demand for total imports is large. When the elasticity of demand for total imports is large, then price discrimination opportunities are small because a decline in sales in any given market will have a relatively small impact on market prices (Booze, Allen and Hamilton, 1995). A further possibility for research in this area is to examine the costs or benefits of price discriminating behaviour (See section 2.2.3).

An alternative assessment, made by McCorriston and MacLaren (2002), compares the producer support estimates (PSE)1 of countries such as Australia and Canada with the USA and the EU, in an attempt to show that the impact of STEs is marginal compared with other trade distortionary programmes. Carter and Wilson, (1999) , agree with the premise that perhaps it is the trade distorting behaviour of the USA and the EU which allows STEs the opportunity to price discriminate within the international market (p 206). Table 2.3 below shows the OECD’s PSEs for all agricultural products and for wheat.

1 PSEs are measures of government support, specifically farmer income supports, transferred from consumers or tax payers.

Table 2.3 Producer Support Estimates

(1998 – 2000)

Country All agricultural goods Wheat Australia 6% 6% Canada 18% 12% USA 23% 45% EU 40% 49% OECD countries 35% 42%

(Source: OECD (2001), table III).

Table 2.3 shows that the PSEs for the USA and the EU are considerably higher than those for all agricultural goods and for wheat in particular, suggesting that the USA and the EU, leaders in denouncing STEs as trade distorting, have programmes that are having a more serious impact on trade than the CWB or the AWB. Regardless of this initial analysis, McCorriston and MacLaren (2002) go on to conclude from their benchmark modelling process that “STEs have the potential to distort trade” with the effect of greater adjustment processes for other subsidising exporters (p 150).

STEs and their potential impact on international trade are not only relevant to the Australian wheat industry but are also topical in agricultural trade negotiations worldwide. There is much scope for further research in this area of agricultural economics.

2.2.3 MODELLING THE WORLD WHEAT MARKET Modelling of the world wheat market has been a concurrent theme in agricultural economics over the last four decades. A framework that accurately depicts the issues of international wheat trade, market power and the role of the STE has not yet been established.

McCalla (1966) was the first to address the issue using an oligopolistic model of the world wheat market, with Canada and the US being the dominant players and Canada acting as a price leader. Alaouze, Watson and Sturgess (1978) completed a study similar to McCalla using a triopoly model which includes the newly emerging exporter, Australia.

McCalla and Schmitz (1979) undertook a qualitative study of the US and Canadian grain marketing systems. They evaluated performance using producer prices and export market shares and government policies and suggest four approaches that can be used to investigate these issues: theoretical constructs; empirical welfare concepts; industrial organisation; and historical, institutional and descriptive approaches. McCalla and Schimtz note the need to adopt a “pragmatic, partial policy-analysis approach to international wheat markets”. Following this, Carter and Schmitz (1979) suggest that world wheat prices are determined by the major importers as opposed to the major exporters (this approach has also been investigated by others including McCalla (1966), and Alaouze, Watson and Sturgess (1978)).

Sarris and Freebairn (1983) model international policies as a Cournot equilibrium interaction of excess demand which are used as solutions to domestic welfare optimization problems. They show that world prices are higher, and price instability lower, under free trade. In the early 1980s European policies represented 80% of these distortions and the US accounted for the remaining 20%. Sarris and Freebairn assume that there is no price leader in the world market, that is, a single country’s policy is not directly influenced by other countries’ policies but only indirectly influenced through world prices.

Karp and McCalla (1983) propose that a dynamic game is preferable to the partial framework as it allows the inclusion of importers and exporters as both have potential market power, in a multi-period framework. A Nash non-cooperative difference game is applied and results suggest that a difference game based on a better econometric model could be useful in policy analysis.

Kolstad and Burris (1986) are the first to note inadequacies of perfectly competitive models applied to international wheat trade. They apply a Cournot-Nash duopsony model to explain trade and to compute the spatial equilibrium in oligopolistic and oligopsonistic markets Kolstad and Burris conclude that the duopoly and triopoly models are better at explaining trade in the world wheat market with the duopoly model giving slightly more accurate results. Paarlberg and Abbott (1986), like Kolstand and Burris, note the need to endogenise policy decisions. They use a revealed preference methodology of interest groups within countries to estimate the policy makers’ conjecture of the excess demand (supply) and then compare this with observed market behaviour. An assumption is made that the objectives of policies are to maximise domestic welfare and also that wheat is a homogenous good, this is a potential limitation.

Pick and Park (1991) use a model based on firm pricing decisions which yields statistical tests of market power, encompassing both perfect and imperfect competition based on industrial organizations theory. Larue and Lapan (1992) take a twist from the game theory approach to market structure and examine country specific reputation mechanism in the pricing of wheat. They use an extension of the Armington Model to examine price premiums and quality differences between a monopsonist and perfectly competitive market. Piggott (1992) introduces the Equilibrium Displacement Modelling (EDM) (comparative statics) to determine market power and gains of a single desk seller, specifically the AWB. Piggott shows that benefits of a single desk may be less than the costs and focuses on EDM as a potential substitute for econometric modelling (often restricted by data).

Just, Schmitz and Zilberman (1979) previously examined this method and found that state marketing boards exercising monopoly and monopsony power can generate greater rents from those that command either state export or import arrangements. Ryan (1994), a former AWB marketing manager, presents a qualitative analysis suggesting the application of a promotion/development scenario (Cost Benefit Analysis) in the US as to how Australia will react without an export monopoly, and notes that Piggott’s (1992) EDM is inadequate as it fails to make assumptions consistent with the structure of the international wheat market

Ahmadi-Esfahani (1994) praises the use of a Cost Benefit Analysis in investigating power in the international wheat market and suggests a plausible model which could serve as a framework for the Single Desk arrangements would be the Sunk Costs Model. This model incorporates exogenous costs include set up costs of establishing a distributional network and endogenous costs such as research and development and advertising costs. Other key components need to include brand names and reputation, loyalty (and market devices), price discrimination and dispersion, quality controls and links to other industries and notes that economies of scale, scope and information can also be captured. Ahmadi-Esfahani criticizes partial equilibrium modeling saying it lacks the capacities required by new trade theory, that is assumes perfect competition, and is therefore highly inappropriate. General equilibrium modeling is more appropriate especially with built in supporting sub models reflecting non- cooperative and dynamic games. It is important to discover alternatives means to estimate various policies and support programmes in an informal/unstructured manner and to look at the impact on trade volumes, this, says Ahmadi-Esfahani, also solves the problem of the lack of adequate data.

…traditional theories, models and estimation methods are incapable of providing effective approaches to the problems facing the Australian wheat industry.

(Ahmadi-Esfahani, 1994)

An alternative to the more traditional market power approach could be the adoption of welfare analyses which examine the benefits and costs of price discrimination such as work by Katz (1984) and Varian (1985). These approaches could be applied to examine the welfare effects of STEs, specifically within agricultural markets, such as the wheat industry.

Katz (1984) examines the issue of whether price discrimination procedures have the ability to improve efficiency and reinforce competitive principals. Using an adapted version of the Salop and Stiglitz (1977) monopolistic competitive model, Katz focuses on second degree price discrimination, the establishment of a pricing structure for a particular good based on the number of units sold, and differentiating consumers as ‘uninformed’ or ‘informed’. Uninformed consumers are assumed to be those who make small purchases choosing from their suppliers at random, whereas informed consumers are the larger purchasers who purchase frequently from the supplier with the lowest price. This implies that the competition between suppliers is far greater when dealing with informed consumers. Katz’s (1984) results show that price discriminatory behaviour is efficient, and hence welfare increasing, under these assumptions provided that there is a larger proportion of uninformed consumers.

…the policy of always allowing price discrimination is more efficient than a policy of always forbidding it.

(Katz, 1984, p 1453).

Although this problem is dependent on the cost and demand conditions. The application of such approaches to the AWB Ltd specifically, or the wheat industry generally depends greatly on the market structure of wheat importers. Love and Murniningtyas (1992) focus on importer power and use the principle of profit maximization and Lerner’s index to test for market power, suggesting that the power of large state trading

24 importers, (in this case, Japan) is substantial. Given that the import market for wheat is heavily dominated by government trade agencies in Japan, Indonesia, China, Egypt, Pakistan, and Tunisia (see section 2.2.2), and where all but Tunisia rank as top importers of Australian wheat (see table 2.2), may suggest that application of Katz’s approach to the welfare effects of the AWB Ltd may not be constructive in this specific case.

Varian (1985) examines the effect of third degree price discrimination, where the firm is able to segment its customers into two or more separate markets where each market defined by unique demand characteristics, on social welfare following from work by Robinson (1933) and Schmalensee (1981). Schmalensee (1981) shows that regardless of whether a price discriminatory approach leads to an increase in output, any increase in output that is associated with price discrimination, will necessarily imply an increase in welfare (Varian, 1985, p 871). This translates as a clear measure of determining that any price discriminatory firm can not be a benefit to society unless output is increasing. Application of such an approach to the AWB Ltd could be an interesting study, although beyond the scope of this thesis.

2.2.4 THE MARKET POWER DEBATE Investigation of previous literature with specific reference to the former Australian Wheat Board allows us to determine the consensus on AWB(I)’s ability to exercise market power. Market power, in terms of trade, on the exporters’ side, can be defined as the ability of an exporter to control or restrict supply in order to command a price premium over and above the world price. Alternatively, it is the ability of an importer to restrict supplies required. Following from the definitions given in the previous section on state trading enterprises, it can be assumed that STEs have the potential to harness market power.

Table 2.4 presents an overview of the literature available on the market power of the AWB’s single desk authority, showing a balance of academic views. Due to the generally qualitative

25 nature of the observed studies a quantitative approach is justified in order to determine whether the AWB(I) does or does not command market power.

Few quantitative and substantial studies have been completed, and there appears to be no formal method of computation. The majority of studies use derivatives of game theory to explain market structure in a qualitative manner. Others have assumed, for the benefit of their research, either that the AWB(I) does or does not command market power, depending on the structure of their models. Several have attempted to compute the AWB(I)’s market power and have achieved inconclusive results due to the lack of appropriate data, such as Paarlberg and Abbott, 1986 (Partial Equilibrium Model with endogenised policy decisions); and Piggott, 1992 (Equilibrium Displacement Model).

Table 2.4 Evidence of the Market Power of the AWB

YES INCONCLUSIVE NO Beard and Purcell, 1996: *Burton and Lobb, 2000: Examine *Wilson and Dahl, 1998: Imply market power through the use the AWB Ltd’s ability to price State Trading Enterprises do not of a Betrand behaviour between discriminate using confidential command market power. AWB and the world market. contract data released for the 2000 NCP Review of the WMA. Ryan, 1994: *Carter et al.,1999: US is in a price Vanzetti, 1991: “The AWB, through internal, leadership position and that Australia has a limited ability to unpublished, unavailable studies, Australia and Canada are price influence world price. confirms market power” (p. 118). taking fringe players. Further research needed. International Policy Council on Watson, 1999: *Kolstad and Burris, 1986: Agriculture and Trade, 1991: Inconclusive – single desks are not US/Canada duopoly more consistent AWB (and CWB) has a price necessary to achieve locational than triopoly with Australia. advantage as a single desk (cited in market power. Ryan, 1994). Miller, 1991: *Piggott, 1992: The AWB has probably extracted Inconclusive – the single desk is not higher average prices in world necessary to achieve the limited markets than otherwise obtainable. market power of the AWB. Alouze, et al., 1978: Industry Commission Report, 1991: Australia/US/Canada triopoly was Doubts AWB’s ability to command the dominant structure of the market power. international wheat market. Freebairn, 1968: Paarlberg and Abbott, 1986: Australia has considerable market Inconclusive – AWB has had market power in sub-markets (i.e. soft power (1969-72) but appears to no wheats) although Canada and USA longer retain it. are best placed to exercise total market power (p 116). *= substantial quantitative study, as opposed to qualitative argument, assumption or unsubstantiated argument.

Data appears to be the biggest limitation in this field, with the only unambiguous and confident result espoused by Ryan (a former Australian Wheat Board marketing manager). Ryan (1984), states that the AWB, “through internal, unpublished, unavailable studies, confirms market power”.

It should be noted that there are other recent studies which focus on the market power of the Canadian Wheat Board. Due to the similarities of the two institutions, it is possible for parallels to be drawn from studies completed on the CWB on both wheat and barley. A study on behalf of the CWB by Kraft, Furtan and Tyrchniewicz (1996) suggests that the CWB could have been commanding premiums of up to C$13.35 per tonne, or C$265 million a year from 1980/81 to 1993/94 as a result of operating a single desk (Kraft et al, 1996). However, these findings proved unsubstantiated as they are not based on premiums received by farmers and the farm gate, and nor do they take into account differences between premiums awarded due to high wheat quality or other services such as flexible credit arrangements, and as such, it is not possible to determine direct causality as to whether these premiums are solely attributable to the single desk (Carter and Loyns, 1996)2.

Carter, MacLaren and Yilmaz (1999) comment that as a result of data restrictions a “direct measurement of market power is not practical” (p 3). They suggest, following work by Goldberg and Knetter (1999), that it is possible to utilise the relationship between Lerner’s Index (p-MC/p) and the inverse demand function faced by the exporting nations to determine if there is a mark up on prices as a result of imperfectly competitive activities.

The Carter – Knetter model is one such approach that could be used to further examine the case of the AWB(I) Ltd, given the appropriate data. The Carter - Knetter price discrimination model is outlined in chapter 3.

2For further information see: Schmitz et al, 1997; Carter and Wilson, 1999; Alston and Gray, 2000; Schmitz and Gray, 2000

2.2.5 CONCLUSION The political economy of the international wheat market is both complex and multidimensional. The number of exporters and importers, their aims and political agendas are diverse. As a result, the issues addressed in section 2.2, such as international trade, state trading enterprises and market power are important considerations to take into account when analysing the impact of the AWB(I)’s position in Australia and internationally.

3 2.3 THE POLITICAL ECONOMY OF THE AUSTRALIAN WHEAT INDUSTRY The Australian wheat industry cannot be easily examined in a purely economic framework. Decades of legislation and Australian government control means that politics and economics are inexplicably linked. The aim of this section is to provide an overview of the political economy of the Australian wheat industry by addressing key historical issues and recent industry events.

2.3.1 HISTORICAL BACKGROUND The Australian Wheat Board was initially created under a Commonwealth government initiative (War Precautions Act) in 1915. This ‘temporary war-time emergency measure’ was designed to ensure compulsory acquisition of wheat from farmers, establish price fixing, control shipping and make advanced payments to growers (Ryan, 1984, p 117).

During the inter-war years a voluntary co-operative and private traders replaced the marketing scheme, however, the Great Depression assured government involvement remained a key feature in the Australian wheat industry. Economic conditions became increasingly dire as the world wheat price plummeted and legislative assistance failed to be procured. The senate passed a price guarantee of 3 shillings per bushel in early 1931, however, due to the unavailable funds, the measure was “still-born” (Hicks and Ireland,

3 For a comprehensive review of microeconomic reform with focus on Australia see Chapter 4.

1997). Finally, in November 1931, the Wheat Bounty (No. 2) Act was passed providing Australian wheat farmers with 4 and ½ pence per bushel for the marketing year 1931/32 (Hicks and Ireland, 1997). Other assistance programmes and debt relief measures were available during the latter part of the 1930s. This was the beginning of a long term trend of federal government support for the industry.

In 1939, following the onset of WWII, the Australian Wheat Board was re-instated (Hicks and Ireland, 1997 and AWB, 1999). There were several advantages to government control in the wheat industry, the most base being that farmers saw themselves as “freed from the dominance of wheat merchants”4 (Hicks and Ireland, 1997).

2.3.2 THE WHEAT INDUSTRY STABILISATION ACT (1948) These attitudes and recognition of possible benefits to farmers led to an appreciation of centralised marketing which resulted in a lobby group for a ‘compulsory national marketing scheme’ following the end of WWII (Ryan, 1984, p 117). The Wheat Industry Stabilisation Act (1948), (WISA), incorporated the four major functions of price control and ‘orderly’ marketing5:

Ø Guaranteed prices; Ø Home consumption price; Ø The ‘official’ establishment of the Australian Wheat Board; and Ø Stabilisation agreements.

(Hicks and Ireland, 1997).

4 Wheat farmers during the early 20th century were being exploited by merchants due to the concurrent technological advances in the storage and transport industry. Merchants often colluded, leading to a decline in prices and persistently high transport costs (Hicks and Ireland, 1997). 5 ‘Orderly’ marketing is defined as the removal of competition between producers and the market (Hicks and Ireland, 1997).

WISA (1948) was enacted with a renewal to take place every five years (Hicks and Ireland, 1997). Between 1948 and 1979 each subsequent Act has retained the basic structure granting the AWB monopsony rights and marketing powers for Australian wheat, the ability to price discriminate on the domestic market, pooling of sales revenue and marketing costs, and government funded ‘buffer’ schemes to assist in the transfer of risk from farmers to the Australian federal government (Hicks and Ireland, 1997).

2.3.3 DEVELOPMENTS IN THE LATE 1970S AND EARLY 1980S During the late 1970s and early 1980s much research was conducted into the impacts of the WISA and other government provisions in the Australian wheat industry. Research was conducted in areas including assessment of the AWB’s “competence and accountability” (Hicks and Ireland, 1997). A scathing report on this specific issue was released by the Senate Standing Committee on Finance and Government Operations in 1979, and reports on the state of the Australian wheat industry were completed by the Industries Assistance Commission (IAC), (1978, 1983).

Recommendations of two main IAC reports (1978, 1983) have been summarised below, and lay the basis for the Wheat Marketing Act (WMA), (1984): Deregulation of the domestic market including removal of import restrictions (apart from reasonable quarantine requirements); Better information to allow assessment of the AWB’s performance in export marketing via publication of separate export and domestic accounts and regular sales by open destination tender;

More efficient price signals and incentives to growers a reduction in cost-pooling by the AWB and BHA (Bulk Handling Authorities), changes in calculation of the first advance, creation of a market in negotiable script for the growers’ remaining equity in each pool, and publication of information on AWB costs of financing the first advance; and

Removal or limitations of some of the powers already granted to the AWB such as that to trade in futures, establish reserves, etc., and including maximisation of returns to growers without taxing domestic consumers.

(Johnston, 1984, p 103).

1979 saw the first of several “drastic changes” made to the existing WISA (Miller and White, 1980). The key changes included the introduction of the Guaranteed Minimum Price Scheme (GMPS) (stabilisation through underwriting)6 and, pricing alterations on the domestic front in order to differentiate prices paid for wheat used in “human consumption” as opposed to wheat used in industrial processes or as feed wheat (Watson, 1984, p109).

2.3.4 THE WHEAT MARKETING ACT (1984) Two important non-political events cemented the changes to WISA, implemented in the Wheat Marketing Act (1984). Firstly, the occurrence of a severe drought in the eastern states (1982) placed huge financial burdens on farmers and the government, as well as substantially decreasing output. Secondly, substantial flooding in the crop year proceeding 1982 contributed to significant degradation of wheat quantity and quality (Hicks and Ireland, 1997).

The creation of the WMA (1984) formalised the following strategic changes (Hicks and Ireland, 1997):

Ø Introduction of a system of permits issued by the AWB for wheat used for stockfeed, this allowed producers and consumers to trade directly under AWB supervision; Ø Alteration to the GMPS (1979) which served to increase ‘market’ signals to producers;

6 The GMPS replaced the WISA developed ‘buffer’ schemes.

Ø Introduction of five distinct categories of wheat for which prices were underwritten to reduce cross subsidisation; and Ø The requirement of two independent industry reports to be undertaken prior to the next amendment (Royal Commission, 1987 and the IAC, 1988). These reports provided the catalysts for future discussions and developments.

2.3.5 AMENDMENTS TO THE WHEAT MARKETING ACT (1989, 1992) The Royal Commission’s report (1987) into grain storage, handling and transport costs was considered a “landmark” inquiry (Hicks and Ireland, 1997) and provided an insight into the efficiency of the Australian grain distribution system. Cost effective measures were outlined and an empirical study was conducted which suggested a national saving of A$10 per tonne could be achieved by devising a strategy for provision of a flexible grain distribution system (Hicks and Ireland, 1997).

A third IAC report (1988) centred on the necessity of continued provision of assistance to the wheat industry and the model used to prescribe any such assistance. The recommended course of action was -

…designed to improve the wheat industry’s competitiveness by removing those regulations which impede growers and buyers of Australian wheat from responding flexibly to market developments.

(IAC, 1988, p 19).

The publication of documents such as the 1987 Royal Commission report and the 1988 IAC report, lead to a period of heated debate “the most controversial in fifty years” during 1988/89. This resulted in legislation that “contained the most significant changes ever made to Australian wheat marketing arrangements” (Hick and Ireland, 1997). The principal changes included the end of domestic pricing arrangements, change in government

33 guarantees, the introduction of the Wheat Indus try Fund (WIF), and the change in the AWB’s objectives. These transformed from a statutory marketing authority maximising farmers’ returns, to having to also take into account minimising storage, handling and transport costs, where costs were to be passed back to farmers wherever possible.

In 1989 the Australian federal government introduced the Wheat Industry Fund (WIF), to release the government from directly guaranteeing AWB’s loans and to allow the AWB a means to finance their own commercial borrowing requirements and to underwrite their investment debt. The WIF was funded through compulsory (2%), levies on all wheat sales up until July 1999, and was operating at A$450 million with 62 thousand equity holders by the end of 1997 (AAFC, 1998). The WIF was disbanded in 1999 when the AWB Ltd “began directly financing pooling and commercial activities” (AWB, 2002).

2.3.6 THE TRANSITION PERIOD (1990-1997) The period from 1990-1997 is central to this thesis. Following the WMA amendments (1989) the domestic market was officially deregulated, and more importantly, the AWB changed its objectives and began operating as a profit maximiser in preparation to become a (semi) private company (1999) under Australian Corporations Law. Restructuring the AWB, in order to ensure minimum disruption, became the industries’ principal goal.

The AWB’s structural change was facilitated through the Grains Council of Australia’s (GCA) “Grains 2000” project which aimed to examine the strategic planning required for the future of the industry (Hicks and Ireland, 1997). Wheat is Australia’s largest and most important grain crop, hence the AWB was an important player in the GCA project developments. It was recognised that the AWB would require a certain degree of agility in order to compete with the developments in the international and domestic grains markets. Discussion at a 1991 CGA conference lead to the establishment of the National Grain Marketing Strategic Planning Unit, to examine key industry issues. Membership included the AWB, GCA, the Australian Grain Marketing Federation, Bulk Handling Authorities of

Australia (BHAA), the National Agricultural Commodities Marketing Association, Australian Malters and Brewers, the Department of Primary Industries and Energy (DPIE), and the Grains Research and Development Corporation (GRDC) (Hicks and Ireland, 1997). Publication of the Australian “Milling Wheat Project” (1995) funded by the GRDC.

The Milling Wheat Project examined the options available to the Australian wheat industry, developed a strategy to focus on marketing and recommended several key points it believed were critical to the industry:

Ø To protect core (export) markets as opposed to the development of new markets; Ø To base competitive advantage on selective differentiation in positioning, selling and handling wheat; Ø To develop ‘defences’ against key competitors, e.g. Canada; Ø To retain the single desk but prepare for competition; Ø To expand the domestic market to include New Zealand and Papua New Guinea; Ø To encourage the AWB to vertically integrate; Ø To permit the AWB to market and export all Australian grains; Ø To corporatise the AWB with vested grower ownership; and Ø To maintain the WIF under AWB control.

(Booz, Allen and Hamilton, 1995, p 56-63).

This report lead the GCA to organise grower meetings across Australia in 1994. These meetings examined possible models of structural design for the Australian wheat industry such as, re-regulation, complete deregulation, maintenance of the current industry structure, corporatisation of AWB with retention of the single desk, and privatisation with the single desk (Hicks and Ireland, 1997). Following these discussions with growers the GCA outlined five key components for inclusion in the AWB restructure:

1. Retention of single desk selling for exports; 2. Retention of grower ownership/control; 3. An adequate capital base to maintain existing levels of harvest payments; 4. Increased commercial flexibility; and 5. Industry self-determination.

(Hicks and Ireland, 1997).

A Working Group7 was established to advise on a suitable corporate and financial framework for the ‘new’ AWB following the above guidelines with recommendations put forward to the Minister for Primary Industries and Energy.

In 1997, the Minister revealed the proposed corporate structure for the AWB Ltd, to be implemented in 1999. The key features are: Corporations Law company under grower ownership and control…responsible for all commercial aspects of wheat marketing; The company will operate as one holding company with two subsidiaries, a wheat pooling/export subsidiary and a commercial subsidiary; and, shares in the holding company will be issued in two classes: A-class shares will be issued to all growers and the WIF will be converted to B-class shares (Anderson, 1997).

7 Appointed by the GCA, AWB and DPIE were independent financial advisers, Bankers Trust, and an independent legal team from Mallesons Stephen Jacques.

2.3.7 THE FORMATION OF AWB LTD, AND IT’S CORPORATE STRUCTURE The dual class or ‘grower corporate model’ (GCM) (figure 2.2) was introduced in line with the privatisation process on 1st July 1999.

Figure 2.2 The ‘Grower Corporate Model’

Class A Shareholders AWB Limited Class B Shareholders (growers) elect 7 Holding company; (investor equity) elect 4 Directors Sole shareholder of AWB(I) Ltd; Directors Responsible for services for other co’s in the AWB Group Obligation to ensure AWB(I) net pool returns to growers are maximized; Service level agreement with AWB(I) and AWB Finance.

AWB (Australia) Ltd AWB Finance Ltd A$5 mil. nominal capital A$5 mil. nominated capital; Responsible for domestic Provides finance for grain trading and growers delivering to commercial activities; AWB(I); AWB (International) Ltd A$200 mil. committed bank Borrows from global capital A$5 mil. in nominal capital; facility. markets; Responsible for export pooling of Service level agreement wheat and maximizing growers net with AWB Ltd and returns; AWB(I). Service level agreement with AWB Ltd and AWB Finance. Pays AWB Ltd a fee for administration, human resources, Agrifood AWB marketing, risk management, Technology Research AWB US AWB Asia funding, shipping and treasury Pty Ltd Pty Ltd A$5 mil; A$5 mil; A$5 mil; A$5 mil; Handles Coordinates services. Handles RandD AWB Sales and quality services. Group marketing assurance activities in in Asia. and testing the USA. services.

Source: Irving et al, 2000, Appendix 4.

The AWB Ltd assumes accountability for all commercial aspects of wheat marketing including financing and wheat pooling arrangements. The GCM comprises of a holding

37 company, AWB Ltd, with two subsidiary companies, AWB (International) Ltd, who is responsible for all export operations, and AWB (Finance) Ltd, which controls all financing issues. Shares are issued in two distinct classes: A–class shares, established for the grower community empowering them to elect a majority of the Board of Directors of the AWB Ltd, and the B-Class share system, issued on the basis of the WIF equity enabling the shareholders to elect four directors of the 13 member Board.

As noted by Watson (1999), there are foreseeable problems with a dual share plan:

There is a conflict of interest between A-class shareholders and B-class shareholders that will be difficult to resolve by the board of the new AWB. Interests of A-class shareholders are served by high pool prices with lower rates of return on B-class shares. The reverse applies to growers with substantial equity in the WIF.

(Watson, 1999, p 450).

This is an important issue for future research into the structural change of the AWB Ltd and its effects on the wheat industry.

8 2.3.8 THE NATIONAL COMPETITION POLICY REVIEW 2000 The Australian National Competition Policy (NCP) Review of the 1989 Wheat Marketing Act (WMA) was announced in April 2000 by the Honourable Warren Truss MP, Federal Minister for Agriculture, Fisheries and Forestry. The aim of the NCP Review was to assess industry behaviour and, where possible, reform legislation which may be restricting competition, “to determine whether there are any net benefits accruing to Australia from the AWB’s wheat export monopoly” (Ireland, 1998).

8 For a more detailed discussion on Australia’s microeconomic reform see chapter 4.

In relation to the WMA Review, guidelines were set up outlining the methods of research, the pro forma for reports, suggested key issues and suggested modelling techniques (Piggott and Edwards, 2000). An independent review committee (IRC) was also established to assess “whether the current legislation provides a net benefit to the Australian community compared with alternatives, and determining preferred options for regulation, if any” (Irving et al., 2000, p 1). There were over 3300 submissions from stakeholders. Of these reports, three are of principal concern, firstly an in-house study by the AWB Ltd., and two external reports, one by the Allen Consulting Group (ACG), and a second by the IRC9. It is important to note that the legislation governed under the WMA is somewhat ambiguous and the IRC notes that a lack of objective, “will make it unnecessarily difficult for any future industry group, forum or review working on wheat marketing arrangements to agree on a common reference point” (Irving et al., 2000, p 1).

The five key issues put forward by Piggott and Edwards (2000) for the NCP Review of the WMA that should be considered are as follows:

1. Price premiums and market power; 2. Domestic impacts; 3. Marketing efficiency; 4. Innovation / dynamics; 5. Implications for liberalisation of global agricultural trade.

(Piggott and Edwards, 2000)

All three of the reports address, in different magnitudes and to different degrees the first three issues, however, the fourth is largely disregarded. Other important issues such as

9 The independent review committee, appointed by the Minister, consisted of Mr Malcolm Irving, Mr Jeff Arney and Professor Bob Lindner.

39 innovation and market development, effects on and efficiencies of trade, have been touched on in various submissions, but with little substance.

THE AWB (INTERNATIONAL) LTD SUBMISSION TO THE NCP REVIEW (2000) There is evidence of some confusion throughout this report including understanding of basic economic theory as well as flawed empirical analysis. The submission is pro the retention of the AWB(I)’s single desk marketing powers based on the principle that the single desk provides “a net benefit to the Australian community and is the only effective means of achieving the objectives of the WMA”10 (p 100).

Market Power: AWB(I) concludes from a misleading empirical study that there is evidence that Australian wheat commands “real premia beyond which that would be explained by distance advantage and particular mix of grades” (p 30). AWB(I)’s confusion as to the economic theory behind market power and price discrimination can be seen in their contradictory admission that Australia is a “price taker” and “cannot restrict supply” (p 25).

Other Benefits: Not surprisingly AWB(I) list copious benefits, above and beyond the ability to command price premia, including “better” risk management for growers, universal rights to access the international market, information advantages leading to efficiencies throughout the supply chain and branding and promotional benefits. Their verification of these benefits are generally qualitative and appear somewhat unsubstantiated. However, given the qualitative nature of the other reports their arguments are as valid as any others presented.

Domestic Markets: AWB(I), although they have addressed the issue of the effects of a single desk on the domestic market, place less weight on this than they perhaps should.

10 The aim of the WMA is threefold and includes the establishment of the Wheat Export Authority to control exports and monitor the AWB(I)’s performance; to grant AWB(I) single desk powers indefinitely; and to continue federal ability to override state and territory governments (Piggott and Edwards, 2000)

They acknowledge that the single desk will “have an impact on domestic wheat prices”, however, they state that “there is little evidence of a negative impact on wheat consumption, nor…(any) adverse affect on the competitiveness of Australian export businesses using wheat as an input” (p 65). Again, little conclusive support for these premises can be found, basically the industry is asked to take at face value that the benefits to Australian wheat growers outweigh any costs to the domestic market (p 65).

Trade Efficiency: “Efficient operations via economies of scale and scope is possible under the single desk”… “work(s) to reduce costs and improve efficiency” (p 57). Again, analysis is of a predominately qualitative nature, and they do not even go so far as to determine the costs to farmers of the current system.

Innovation and Market Development: As commented on previously, the submissions reviewed generally failed to investigate the issues of innovation and market developme nt under the single desk. AWB(I)’s section 6.3 heading sums up their views, “The Single Desk already provides a clear vision for the future” (p 88).

THE ALLEN CONSULTING GROUP SUBMISSION TO THE NCP REVIEW (2000) This report provides a good mix of qualitative and quantitative analysis of the issues at hand. However, although ACG employs an empirical investigation it should be noted that they, like the AWB(I) failed to take Piggott and Edward’s (2000) modelling recommendations. There is also some skepticism, as with AWB(I) analysis, on the validity of the modelling methods used and the results obtained. It should be noted that ACG finds the AWB(I) quantitative study to be “overstated and largely implausible” (p 34).

Market Power: ACG denies any presence of market power by AWB(I) and uses a reasonable “first principles” argument to sustain their argument. “The existence of low entry barriers and AWB’s (sic) ability experience (sic) as a price taker…suggests that the single desk is ineffective in providing AWB with market power” (p 19). AWB(I)’s ability to

41 command price premia is not refuted, however, ACG remains sceptical of the origin of the premia and notes that “price premiums do not necessarily indicate market power, but can reflect differences in product quality, reliability of service and relationships (with consumers)” (p 19).

Other Benefits: The report, at this time (see footnote 2), fails to adequately address the other benefits of a single desk selling arrangement.

Domestic Market: ACG take a more conservative view than either of the other reports commenting that “the precise degree of this consumer impost is unclear” (p 42). It is important to acknowledge that all reports are in accordance with the fact that the existence of a single desk export arrangement will have some degree of negative impact on domestic wheat prices. Basically, further research using consumer and producer surplus theory is required for anything to be conclusive.

Trade Efficiency: Again ACG fails to make any conclusive comments and unfortunately has not attempted to calculate the costs to growers of a single desk. They state, as a result of qualitative analysis and results from previous studies, that “the dynamic benefits of competition are generally significant” (p 56).

Innovation and Market Development: Taking a comparative approach, using the example of the deregulation of the South African deciduous fruit export single desk, ACG concludes that, through a voluntary levy arrangement there would be “increased expenditure on research and development” (p 32). Basically, ACG conclude that innovation and market development can occur regardless of market structure.

THE REPORT OF THE INDEPENDENT REVIEW COMMITTEE TO THE NCP REVIEW (2000) There were several key recommendations focusing on enhancing competition in the Australian wheat industry that were put forward by the IRC. These included: a clear

42 definition and specification of the legislative objectives of the WMA; the total independence of the WEA as well as licensing system to replace the permit system operated by the WEA to enable farmers to self-export by bag and container; a scheduled review of the AWB Ltd by the WEA in 2004; the establishment of a joint industry-government forum; and, the retention of the ‘single desk’ (IRC, 2000).

The IRC submitted their final report to the Minister in December 2000 with the underlying premise that the single desk be retained. The Minister’s response to the committee’s recommendation of the retention of the single desk arrangements till 2004 was positive: “The single desk was established in the interests of Australian wheat growers. With around $3.5 billion worth of wheat exported every year, it’s vital that its integrity is protected” (The Hon. Warren Truss, MP, 2000).

It is important to note, that with respect to the recommendation of retention of the single desk, the IRC’s decision was made despite the lack of substantial evidence to conclude if net social benefits of the single desk existed.

The estimation of net benefits is a complex and difficult exercise…Despite some claims that substantial premiums are being earned…considerable evidence was provided that the ‘single desk’ has had an anti-competitive effect on the grain supply chain.

(IRC, 2000, p 144).

The NCP Review of the WMA has highlighted the need for further analysis on the AWB(I)’s ability to price discriminate as a result of the single desk legislation which remains in place. The WEA has scheduled a review of the AWB Ltd in 2004 and the lack of data and study in relation to the AWB Ltd indicates that research should begin now and should take into account the mistakes made in the NCP Review process to ensure that more robust evidence can be found to determine whether there exists a net public benefit as a

43 result of the single desk exporting arrangements held by the AWB(I) under Australian federal government regulation.

2.4 THE FUTURE OF THE AUSTRALIAN WHEAT INDUSTRY Since the conclusion of the 2000 NCP Review of the WMA, preparations have been underway to determine the benefits and costs of a single desk operation for the export of wheat on the Australian community. The WEA Review of the single desk is set for 2004 and independent reports (Accenture, for the WEA, 2002), have already been published outlining the aims of the review process.

Apart from this legislative component, other key industry issues require the attention of researchers. Firstly, following from this thesis, it is important that the behaviour of the AWB Ltd be monitored to investigate the impact of structural change on the industry. Future research could focus on the political and economic ramifications of the AWB Ltd’s aims of simultaneously maximising farmers’ and shareholders’ returns. Secondly, a medium to long term issue relating to the single desk, is compliance to the World Trade Organisation’s (WTO) agreements concerning state trading arrangements (Ryan, 1994 and Hicks and Ireland, 1997). Australia’s statutory marketing arrangements are permitted under the WTO (1997), however the United States has shown interest in pursuing the issue of state trading arrangements at future WTO forums (for reasons given above in section 2.2.2) (Hicks and Ireland, 1997). Ryan (1994) focuses on the opportunities for firms such as the AWB Ltd to thrive in an environment which aims to cut tariffs and export subsidies, “however, it is believed to be beneficial to Australian grain growers through improved prices, greater market access and an increase in demand for grain and grain based products resulting from the expected general increase in world income” (Ryan, 1994, p 108).

2.5 CONCLUSION Chapter 2 has qualitatively examined the international wheat market, the Australian wheat market and the interaction of the AWB Ltd in both spheres. Specific attention has been

44 applied to STEs, their potential adverse effects on trade, and the ability of the AWB(I) to command market power. An historical overview of the Australian government policies and programmes that have affected the Australian wheat industry, including focus on the recent NCP Review, is also presented. This provides a background and context to the issues addressed in the remainder of this thesis. In the next chapter, a price discrimination model is adapted, as a consequence of the outcomes of the NCP Review process, and data analysed in order to determine the AWB(I)’s ability to command market power.

CHAPTER 3 A TRADITIONAL ANALYSIS OF A PRICE DISCRIMINATING MONOPOLIST

3.1 INTRODUCTION This chapter11 presents a traditional analysis pertaining to the AWB(I)’s, ability to act as a price discriminating monopolist in the international wheat market. Price premiums and market power are of paramount importance for a STE such as the AWB(I) and if it could be shown that substantial and consistent price premiums exist, and can be attributed to the actions of the AWB(I), then a central plank in the argument for retention of the single desk regulations would be made.

Firstly, sections 3.2 and 3.3 provide a background for this chapter focusing on the modeling approaches. Section 3.2 examines the basic price discrimination model used in the ACG Report for the 2000 NCP Review. A review of a more complex price discrimination approach to examining market power, the Carter-Knetter Model, is presented in Section 3.3. Following from the discussion on the 2000 Australian National Competition Policy (NCP) Review process in Chapter 2, the possibility of the AWB(I), earning systematic price premiums is investigated using a theoretical and a statistical framework in sections 3.4 and 3.5. Section 3.4 examines the sensitivity of the equilibrium simulation model to changes in the assumption of the functional form of the demand curves12. Section 3.5 looks at an estimation of pricing to market models for the different classes of wheat on disaggregated country data, including an identification of any response in pricing to exchange rate variations. The equilibrium simulation model is then re-solved using only estimates of price

11 This chapter is based on work undertaken by Burton and Lobb (2000) for the Independent Review Committee for the NCP Review of the Wheat Marketing Act. 12 This extends the models presented in the Allen Consulting Group report prepared for the 2000 NCP Review of the WMA (1989).

46 differentials which are statistically significant, and robust, and conclusions are drawn in section 3.6.

3.2 ACG METHODOLOGY Following from the summary in chapter 2, this section will provide an indepth discussion on the methodology used to calculate price premiums by the Allen Consulting Group (ACG) in Appendix B to their report for the 2000 NCP Review of the WMA. It is important to note that the use of this basic price discrimination model (Carter, 1993) is seen as somewhat of an “industry standard” amongst agricultural analysts (notably in wheat and barley markets) and other participants attempting to obtain price premiums (See MSG, 1996; CIE, 1997; Gropp et al., 2000). As a result, it is important that the limitations of such a model are identified, and where possible, alternative modeling procedures that overcome these limitations are developed.

Using a standard profit maximizing, price discrimination model, ACG assumed a Cobb- Douglas demand function (equation 3.1 below) for Australian wheat in each of its, M, markets assuming market power, perfect information and that, within grades, wheat is a homogenous good (ACG, 2000, p 71).

M b i Qi = åaiPi for i = 1,..., M …(3.1) i=1

Where all parameters are market specific and where b represents the price elasticity of demand.

The monopoly sets prices equating marginal revenues across all markets (given perfect information) so as to achieve the following condition:

1+ 1 P b i = j …(3.2) Pj 1+ 1 bi

Using the AWB Ltd’s contract data for prices (P) and quantities (Q) for all markets, values 13 for the other parameters were calculated, assuming one elasticity, b1 = -8, to be given .

Assuming the average return (or average actual price, Ppd) to Australian wheat farmers as the weighted average of all prices received to all the relative quantities sold (wi):

M Ppd = åwi Pi …(3.3) i=1 then, the price premium, PREM, is the difference between the average actual price received

(Ppd), and the perfectly competitive price (Pc), defined as the equilibrium price under perfect competition where price elasticity of supply is zero (equation 3.1). Following from this, the price premium is defined as:

M å Qi Pi PREM = i=1 - P …(3.4) Q c

The price premium was then calculated for various price elasticities of demand (mean of –8 and variance of 4 for a normal distribution), and maximum price premiums were found over this range.

13 The ACG Report used a price elasticity of demand of –8, based on estimates form the MONASH model, the Murphy general equilibrium model and GTAP (ACG Appendix B, 2000, p 75-76). This elasticity is adopted in the Burton Lobb approach later in this chapter.

The data used by the ACG was the same AWB Ltd’s individual contract data used in the analysis below (see section 3.4.2). The ACG calculated a weighted average price for each market for four aggregated wheat types (ACG, 2000, p 75). This data aggregation was one of the major limitations of the ACG report.

The ACG did recognize that their model was also limited by the assumption of perfect information. Uncertainty was placed on the price elasticity of demand, as opposed to uncertainty on the demand function. ACG used a Monte Carlo simulation on the range of the price elasticity of demand parameter, b. Results concluded that under ‘uncertainty’ the range of premiums was A$0.60 with a mean of $1.31 (ACG, 2000, p89-90).

3.3 THE CARTER – KNETTER PRICE DISCRIMINATION MODEL

3.3.1 INTRODUCTION The Carter-Knetter model, developed in Carter et al (1999), is examined in detail as it lays the basis for the modeling approach used in section 3.5 to determine the AWB(I)’s ability to price discriminate. This method appears to be the most straightforward approach to the investigation of market power in the literature, especially given data restrictions which mean a direct measurement of market power is not practical.

3.3.2 REVIEW OF CARTER ET AL (1999) Carter et al (1999), investigate the issues surrounding the suggestion of world wheat trade being imperfectly competitive. Imperfect competition is important for two reasons: firstly, the suggestion of possible strategic ends for government intervention in imperfectly competitive international markets (based on New Trade Theory); and, secondly, the importance of, and implications for, state trading enterprises at the next round of WTO negotiations.

Focus on the existence of imperfect competition in grain trade follows from positive results (i.e. existence of imperfect competition on either the exporter or importer side), by Kolstad

49 and Burris (1986), Karp and Perloff (1989) and Love and Murniningtyas (1992). Carter et al find these results “surprising”:

How is it that price can be set different from marginal cost for a commodity produced around the world with ease of entry unless there are increasing returns to scale?

Carter et al (1999), focus on generic wheat exports to Japan from Australia, Canada and the US. To determine the existence of price discrimination the Japanese import market is used as a basis for comparison as all three exporters have a high market share in Japan. Carter et al (1999), note that Canada and Australia control exports through single desk arrangements and Japan has control over wheat imports through a government importing agency.

Imperfect competition is measured by calculating the elasticity of the residual demand for Japanese imports for the US, Canada and Australian exports, this then allows for any markup over marginal cost to be seen. Hence, following work by Goldberg and Knetter (1999), it is possible to utilise the relationship between Lerner’s Index (p-MC/p) and the inverse demand function faced by the exporting nations to determine if there is a mark up on prices as a result of imperfectly competitive activities. This residual inverse demand equation has price (the unit value of wheat exported to a region or country), as a function of the quantity of wheat exported to that country, a vector of cost shifting variables (input prices and exchange rates) for each export competitor and demand shift variables (real income) in the importing nation. The elasticity of the residual inverse demand is given by the coefficient on the quantity variable.

u u u c c a a ln pt = a + h ln Qt + b ln Wt + b ln Wt + g ln Z t + e t ….(3.5)

Where: P = price Q = quantity Z = demand shift in importing country h = elasticity of residual inverse demand b = coefficient of vector of cost shifters in competing country g = coefficient of vector of demand shifters in importing country t = time e = error u = USA a = Australia c = Canada

If the exporting country has no market power then changes in quantities exported will not alter prices and the residual inverse demand function will be horizontal. If the exporter’s price is determined by shifts in competitors’ costs and not by the amount they are exporting then the exporter has no market power. However, if there is a negative relationship between quantity exported and price received then there will be market power. In other words, there are systematic issues such as quality and the provision of services and distortions from within the importing country (e.g. import quotas), which may affect estimates.

3.3.3 RESULTS Carter et al (1999) apply a two stage least squares approach to estimate the above equation (3.5), giving the initial results.

1 = 0.93 The coefficient on the residual inverse demand elasticity ( e ) for the US, is correctly signed and is, in absolute terms, significantly different from zero, this shows the markup over marginal cost. “Thus, the conjecture that the US has a horizontal residual inverse demand function is rejected and with it, the conjecture of competitive behaviour by the US”.

No significant markup over marginal cost is found for Canada. This is consistent with Canada having a horizontal residual demand function (hence a price taker), and Carter et al (1999), indicate that this implies the Canadian export price does not vary with export volume. It is suggested that any price variations are a result of changes in competitors’ costs and shifts in Japanese import demand. The Australian coefficient is correctly signed but is not significantly different from zero. This implies there is no markup over marginal cost for Australian wheat.

These results suggest, in direct relation to the Lerner Index, that the US is a possible price leader in the Japanese market and Canada and Australia are situated on the “competitive fringe”. Alternatively, Carter et al do acknowledge that there is a possibility that Japan has buying power (“monopsony”), relating to a study by Love and Murniningtyas, 1992. Therefore, three different scenarios are focused on in the application to market structure, competitive pricing; US price leadership with a competitive fringe; and, monopsony (Japanese buying power).

It is interesting to note that instead of a two stage least squares approach it may be preferable to use seemingly unrelated regression as a more appropriate method of estimation given the interaction in the market place. In other words Japan will (in reality), import wheat given information on all three exporters, simultaneously.

Following these preliminary results a “structural econometric model” is derived and estimated for the six endogenous variables (price and quantity for Australia, Canada and the US). Each model is “nested in a general linear model through the use of cross-equation restrictions”.

The Competitive Pricing Model stands as follows, for the ith exporting country:

pi = g i0 - d iiqi + å j d ij p j + g i1 y, i = u,c,a, i ¹ j …(3. 6) where:

pi = the price in country i’s exports in the importer’s currency qi = the quantity exported by country i y = the total expenditure on wheat imported by Japan. Supply function:

pi = q i0 + q i1qi + qi 2 PP i …(3.7)

With PP i being a proxy for input prices in the exporting country i, calculated in Japanese Yen. This proxy can be viewed as the opportunity cost of producing alternative goods. In the US, the proxy price is for corn, in Canada, for canola and in Australia for the production of wool. Carter et al., do not explain the rationale behind the choice of these goods, but it can be assumed that farmers that grow wheat would grow one of corn, canola and wool if they were likely to earn more, respectively. Profit is then maximized (total revenue less total costs), for export sales, disregarding the domestic market, giving the following first order condition:

pi = q i0 + (d ii + q i1 )qi + qi 2 PPi …(3.8)

Equations (3.6) and (3.8) are then expanded for each exporter and form a set of six simultaneous equations.

The Monopsony Model, with all exporters being price takers, shows Japan’s marginal revenue function for imports from the ith exporter:

MRi = g i0 - 2d iiqi + å j d ij p j + g i1 y, i = u,c, a, i ¹ j …(3.9)

With the average outlay function for wheat from each exporter being the same as equation (3.3). The first order condition, given profit maximization is:

pi = g i 0 - (2d ii + q i1 )qi + å j d ij p j + g i1 y, i = u,c,a, i ¹ j …(3.10)

Equation (3.7) and (3.10) are then estimated.

With the US Price Leadership Model, the inverse demand function faced by exporters is defined by equation (3.6), with per unit costs given by, ci = q i PPi , where each exporter again maximizes profit under US leadership, with TC i = qi PPi qi .

Hence the first order conditions for the US will be:

pu =d uub1+ducd cu +d uad au gqu -qu PPu …(3.11) and for, Australia and Canada:

p j = q j PPj , j = c, a …(3.12) Here, equations (3.6), (3.11) and (3.12) form the system of simultaneous equations.

Vuong’s (1989), test for non-nested models is used to determine which market structure is best represented by the statistical data. Firstly, demand (and supply) equations are estimated and then the first order conditions, by the full information maximum likelihood (FIML) method. Carter et al., do not report the FIML estimates as: “the magnitude and the signs do

54 not contribute to our analysis”. The second stage of the test calculates the likelihood ratio

( L f - Lg ) for each comparison ( M f , M g ) and is then normalized by:

1 ^ n ^ ^ -1 ^ ^ ^ -1 ^ 1 2 1 ¢ ¢ 2 2 n wn = [å cu ft å f u ft -u gt åg ugt h ] …(3.13) 2 t=1

The null hypothesis is set as assuming that the data fits all models well. The results show that the competitive and monopsony models are significantly better than the US price leadership model, however, there is no significant difference between the competitive and monopsony models.

Carter et al., conclude that further research is needed to confirm the structure of the international wheat market. This study suggests that imports into Japan are imperfectly competitive on the export side and secondly, that perhaps the US is in a price leadership position and that Australia and Canada are price taking fringe players. However, it should be noted that the results generated with respect to the fit of the data to a US price leadership model were not conclusive, “no compelling evidence of imperfect competition on the exporters’ side” and “Overall, our findings suggest that we cannot rule out the competitive model” (Carter et al, 1999).

3.4 FUNCTIONAL FORM OF THE DEMAND CURVE In this section a number of issues are addressed relating to the sensitivity of the estimated results derived from an equilibrium trade model used by the ACG in their review of the WMA for the NCP (2000). The ACG have assumed constant elasticity demand curves for a profit maximizing price discriminating monopolist. This in itself is not an unusual practice as constant elasticity demand curves are commonly used in modelling processes. On analysis of the ACG’s results it was apparent that different results could be obtained by using linear demand curves. This begged a detailed analysis into the importance of functional form in empirical determination of a firm’s market power.

There is specific focus on the implications of using a linear demand model as opposed to the more conventional constant elasticity demand model. It transpires that the results using a linear demand model generate a consistently higher estimate of premiums.

3.4.1 THE LINEAR DEMAND MODEL FOR PRICE DISCRIMINATION Following from the model used by the ACG14, a price discriminating monopolist will set relative prices such that:

P 1+1/e i = j for i,j=1,...n …(3.15) Pj 1+1/e i

Where i,j are country identifiers, and the e's are the elasticities of demand in the various countries. Once the appropriate relative prices have been identified the overall level of prices needs to be set such that the aggregate demand across all n markets equals the fixed supply that is available for each year.

This formula holds for all demand curves. In the case of the constant elasticity demand curve the e's are directly the coefficients of the demand curve:

e i Qi = ai P …(3.16)

In the case of a linear demand curve, this is slightly more complex, where:

Qi = ai + bi Pi …(3.17)

14 This model can be found in Appendix B of the ACG Report to the NCP Review of the WMA (2000)

With the elasticity given by:

ei = bi Pi / Qi …(3.18)

Assuming the grain seller knows the coefficients of the demand curves, then solving for the optimal price ratio leads to:

P 1+ Q / P b i = j j j Pj 1 + Qi / Pi bi or

a jbi - aib j Pi - Pj = …(3.19) 2bib j

Where the optimal price rule is now in terms of price differences, and depends on both parameters (a and b) of the demand curves.

Identifying the counter-factual market clearing price, for a linear demand function the analysis proceeds in a similar manner to that used for the constant elasticity model. The observed relative prices are used to infer the parameters of the underlying demand curves and to identify the single price, which equates aggregate demand with fixed supply. As with the constant elasticity model, one elasticity needs to be imposed and from this all other inferred elasticities can be identified15. Assuming price discrimination and using the observed prices and quantities the parameters of the underlying demand function are inferred. Thus:

15 The choice of -8 as the baseline elasticity is somewhat arbitrary, but is the mean value used in the ACG Report.

P 1+ Q / P b i = j j j …(3.20) Pj 1+1/(-8)

From equation (3.20), bj can be identified and hence the value of aj identified by:

a j = Q j -bjPj …(3.21)

Hence all parameters of the demand curves are identified. A single equilibrium price can then be found which allocates all of the available supply across the markets. This price is then considered to be the counter-factual free market price, and is used as a basis for identifying premia being earned16.

3.4.2 UNCERTAIN DEMAND FUNCTIONS AND THE ASSUMPTION OF PERFECT INFORMATION The question for a price discriminating monopolist, such as the AWB(I) Ltd, is how to maximize profit (or returns), by allocation of a fixed supply of wheat across countries, that is equating marginal revenues across all buyers. This is easily done provided the supplier has knowledge of the demand curves of their consumers.

The concept of imperfect information and uncertain demand schedules is a highly complex area and opens the door to further research. Suggestions include determining if this uncertainty can be modeled empirically – that is, with time series data is it possible to assume that as the volume of trade increases to a specific market, will uncertainty about the nation’s demand decline (i.e. is lagged market share be a determinant of current prices). Other extensions could include the application to a constant elasticity of demand system, or

16 Note that because of the linear function form of the demand curve it is possible for some countries to have zero demand at the equilibrium price.

58 a situation where the supply is fixed (for example by a quota system), and the demand curve is thus kinked at this point.

It is assumed in this study, for simplicity that, demand curves exist under perfect information, although it is important to acknowledge that if imperfect information were to be taken into account it is possible that the implications of this work could be quite different.

3.4.3 DATA The data set used for this study was released by the AWB Ltd and consists of over 2000 individual contracts of wheat of all classes17, exported to all countries over the period 1997- 199918. For this specific analysis individual contract data was used as opposed to an aggregated form. Although, in principle, this is the most appropriate method for data use, some issues were raised by this analysis and these need to be noted.

Firstly, for some classes of wheat and some countries very few contracts were traded. It is unlikely that robust country specific effects could be identified with these small amounts of information. This is resolved by only including countries with ten or more contracts traded. Ten is an arbitrary number, but it is important that the estimates of price premia should be based on some systematic variations in prices.

Secondly, it does not seem to be appropriate to aggregate data into regional groups, as was done in other reports (CIE, 1997), even if this would overcome the lack of data relating to the first issue. The crux of the pricing to market concept is the idea that some inelasticity in demand for the product can be exploited, and premia earned. Given the reliance on the idea of developing niche markets and building client relationships in the argument for why

17 Wheat class is defined as the different quality receival standards set by the AWB Ltd, for example, Prime Hard, Australian Standard White. See Appendix 3 for a full discussion on the quality and classes of Australian wheat. 18 Due to the commercial nature of this data detailed results cannot be presented in this thesis.

Australian wheat should command a premium, it is difficult to see why aggregate regional groupings would be appropriate.

Finally, six classes of wheat are investigated separately 19 as opposed to being aggregated across classes, as it is possible that within a country a niche market may be identified for a particular class of wheat and premiums may be earned. Alternatively in another class of wheat there may be no price discriminatory behaviour taking place.

20 3.4.4 RESULTS Table 3.1 below reports the estimated average premium for class (1) of wheat over three years (1997-1999), for linear (L) and constant (C) elasticity demand functions. These estimates have been derived using the average prices and total quantities imported by those countries with more than 10 contracts in this class over 3 years. This ensures that the prices and quantities are based on significant volumes of trade (and conforms to the later statistical analysis: see section 3.5). These results are estimated on a different basis to those of the Allen Consulting Group, and cannot be compared with them.

Table 3.1 represents a sensitivity analysis across two aspects of the model: functional form and the choice of baseline country, with an elasticity of demand of -8.

19 Feed wheat was excluded from the data set used as there were not enough individual contracts for the export of feed wheat. 20 See Appendix 2 for the detailed regression analysis output.

Table 3.1 Simulated average price premia derived across all countries

Assuming constant (C) and linear (L) demand curves, for different 'baseline' countries (X, S, T), class 1 wheat

Country where elasticity =-8 1997 1998 1999 X C 1.79 3.23 2.04 L 3.53 6.61 4.17

S C 2.02 3.58 2.29 L 3.82 7.29 4.50

T C 2.02 2.36 1.95 L 3.81 5.12 4.03

The results are the average estimated premia across all grades for class 1 wheat in each of the reported years. The first row reports the values when country X is selected as a country with the elasticity of demand of –8, and the elasticity for the other countries are inferred from the relative prices. The model is then re-solved for two other countries, having an elasticity of -8.

Table 3.2 reports the same exercise for an alternative class of wheat (class 2).

Table 3.2 Simulated price premia

Assuming constant (C) and linear (L) demand curves, if X is the 'baseline' country (class 2 wheat)

Country where elasticity =-8 1997 1998 1999 S C 1.22 1.06 1.07 L 2.44 2.28 2.99

P C 1.40 1.22 1.22 L 2.68 2.55 3.42

W C 1.87 1.17 1.17 L 3.28 2.46 2.46

Y C 1.43 1.74 1.64 L 2.72 3.35 4.79

Z C 3.67 3.29 1.81 L 5.52 5.59 3.77

The results in both tables suggests an important point, that the use of the linear demand equation consistently leads to higher simulated premia, of the order of 2 to 3 times. A further feature of the model is the change in the implied elasticities of demand in each country across the years. Thus, as the observed relative prices change across the years, the implication is that the relative elasticity of demand must also be changing.

For example, Table 3.3 reports the estimates for the class 2 wheat case, when country S (the lowest price country in all three years) is selected as the baseline country with an elasticity of -8.

Table 3.3 Imputed elasticities of demand prices

Actual relative price, class 2 wheat

Country S P Z W Y 1997 -8 -6.94 -4.28 -5.65 -6.80 1998 -8 -6.86 -4.41 -7.16 -5.44 1999 -8 -7.02 -4.30 -7.48 -7.02

For some countries there is relative stability (e.g. country P), but note the variability in the estimate for country W and, country Y. Not only are there significant changes in the elasticity across the years, but the relative sizes of the elasticities also vary.

3.4.5 CONCLUSION As the relative price differentials depend on the relative elasticities, the implications of section 3.4 is that the AWB(I) is assumed to have both identified the shift in elasticities in these years for these countries, and then has changed the relative price levels in response. Key results are that the size of the price premia is very dependent on the functional form and also that elasticities are changing across time. This leads to a major criticism of the simulation model, and consequently biased results, in the report prepared by the ACG (and other work), in relation to the ability of the AWB(I) to price discriminate.

3.5 SYSTEMATIC PRICE PREMIUMS A common assumption suggests that all observed variation of prices are due to a monopolist’s ability to price discriminate through exploiting the different elasticities of demand in the market place. As noted by MacAulay (2000), “Since it is possible that factors other than the ability to price discriminate may cause price differences between countries from a single desk seller then it is difficult to attribute direct causality”. Hence it is important to determine if price premiums are systematic or simply driven by shifts in other factors, such as:

Ø Seasonal timing of contracts; Ø Quality within a class; Ø Country specific effects; or Ø Pure noise (some random variation not explicable by theory).

If the hypothesis is that all elasticities of demand are the same, and there is no possibility of exploiting market power, and yet prices contain some random variation, the simulation model will mechanically derive estimates of premia based purely upon random variation. In order to determine the degree of price variation due to the single desk structure of the AWB(I), ideally one would chose to base the simulation models only on the systematic elements of the price differentials. This requires a two stage process using a statistical model, which is outlined in the following sections.

3.5.1 THE APPLICATION OF THE CARTER-KNETTER MODEL The Carter/Knetter model has been adapted from work developed by Carter et al (1999) and is used to identify the systematic components of the price differentials (See section 3.3 above):

ln( FOBit ) = bi + ct + ei t for all i=1…,n …(3.22)

The log of the free on board (fob) price of a particular grade of wheat to country i , in time period t, is given by the function of a country specific intercept, bi, that is, different countries could be charged different prices, and of ct, a sequence of coefficients for each time period, that strip out any generic affects that are shifting the fob prices period to period and exist for time periods but are not country specific, plus some random error term.

When elasticities of demand are equal then there can be no possibility of exploiting market power. Hence, when bi=bj=0, expected fob prices will be equal. If there is a difference between these expected fob prices then this variation cannot be attributed to market power and must be attributed to some other factor/s. By assuming that all price variation can be accounted for by price discrimination suggests that the model is incorrectly specified. Hence when the simulation model is run it will automatically derive the price premiums based on random variation.

The inclusion of a random error term, eit, in this model suggests that observed prices will not be the same as these differences cannot be linked to individual countries or time. If some of the b’s are significantly different, that is country specific effects exist, then price discrimination is occurring.

Beyond the constant elasticity case, the analysis becomes more complex. For example, the empirical specification used by Carter includes the exchange rate for each country. The exchange rate is introduced in equation (3.23) with ER, the country specific exchange rate, bi and di are other country specific effects and ct, is a time specific intercept.

ln( FOBit ) = bi +ct + di ln( ERit ) + eit for i=1…,n …(3.23)

Where the elasticity of demand across countries is not equal, the optimal fob price to a specific country, i, will vary as the exchange rate varies. Therefore, if there is statistical

65 significance of these elasticities being unequal, there is evidence to support market power.

Having a significant di, indicates that fob prices are responding to changes in exchange rates between countries. Given a perfectly competitive market and the law of one price, this should not be happening. If the demand function is linear, as you change ER this changes the optimal prices, but if the demand function has a constant elasticity of demand curve then as ER changes there is no effect (See section 3.5.3).

It should be noted that the double log functional form of this model is not consistent with any specific demand function. For Carter et al (1999) this is not an issue, as they do not identify a pricing to market effect and therefore do not need to determine the underlying demand function. In a general case, this issue needs to be addressed in more detail. The identification of the existence of market power is not enough, the counterfactual free market price, which determines the degree of price premia that is earned also needs to be recognized. This requires the underlying demand curves to be specified.

In order to identify the underlying demand functions a base country needs to be chosen, country Y. Expected price differentials between country Y and other countries can then be determined by simulation. Equation (3.24) shows the log of fob price in country Y and time period t as a function of b’Y, the estimated country specific coefficient for Y and d’Y is the estimated exchange rate effect, and c’t is the time effect. In the empirical work that follows it is important to note that the b’s and d’s used are only those that are statistically significant.

ln( FOBYt ) = b'Y +c't +d 'Y ln( ERYt ) …(3.24)

This equation is also specified in the same way for country i, which gives the following price ratio:

FOB it bi -bY di -dY = e ERit ERYt …(3.25) FOBYt

If the country specific effects are not statistically different to those of country Y, then the first term cancels (b’i - b’Y=0). If both exchange rate coefficients are insignificant then the predicted price will be equal (FOBY = FOBi).

This process of estimating relative prices was completed by comparing every country against country Y for each of the six classes of wheat using only country specific effects significant at the 10% level. The price used for county Y was the weighted average price for each class of wheat in each year, and all other prices were simulated prices based on the estimated equation (equation (3.25)). These relative prices were then used to calibrate an equilibrium market model assuming constant elasticity demand curves.

3.5.2 THE REVISED EQUILIBRIUM MODEL The statistical analysis in section 3.5.1 suggests there are a number of systematic, country specific effects which can be identified in the contract data for most classes of wheat. It is important to determine how these effects can be transformed into an estimate of the aggregate premium for each class of wheat.

The method used is to simulate expected price differential between countries. As country Y trades in all classes of wheat it is convenient to use country Y as the baseline country.

The predicted (log) price for a wheat of a specific class, sold to country Y is given by:

' ' ' ln( FOBYt ) = bY + ct Dt + dY ln( ERYt ) …(3.26)

Where b'Y is the estimated country specific coefficient for country Y, and d'Y the estimated exchange rate effect for country Y.

The price to country i will be given by:

' ' ' ln( FOBit ) = bi + ct Dt + di ln( ERit ) …(3.27)

Hence, the price ratio:

' ' ' ' FOBit FOBYt = e(bi -bY + di ln( ERit ) -dY ln( ERYt )) …(3.28)

Note that the time dummy variables cancel out, as they are common to all countries. If the country specific effects are not significantly different from that of country Y then the first term equates to zero (i.e. b'i -b'Y =0), and if both exchange rate coefficients are insignificant then the predicted prices will be the same. This is the case, for example, for the class 2 price for country P: the country P specific dummy is not significant, and neither country P’s nor country Y’s exchange rate effects are significant, therefore the model predicts that the fob price for country P and country Y will be the same.

When there are no exchange rate effects, but there is a significant country specific dummy, then this country specific dummy will determine the relationship between the two prices. For example, in the case of class 2 wheat, for country Y and country Z it is implied that the price paid by country Z is some 11% higher than that for country Y. Alternatively, when the exchange rate coefficients are significant, the n the relativity between the two fob prices will depend upon the exchange rates for each country. In this case an assumption is made about the time period (t), (as the exchange rates vary). In the simulation results reported below, a calendar year average of daily exchange rates is used.

For each country, within each class of wheat, an estimate is made of the relative price differential as compared to the base country, country Y, using only those country specific effects, which are significant at the 10% level.

These estimated price relativities are then used to calibrate an equilibrium market model, assuming constant elasticity demand curves. Therefore, regardless of the estimation of an ad hoc price equation, in the simulation it is assumed a constant elasticity functional form approximates the demand curve.

The price in country Y is the weighted average price for the class of wheat in that year, and all other prices are the simulated prices, based on the estimated equation. Therefore those countries where there are no significant effects are assumed to have paid the same price as country Y for their wheat, and hence have the same elasticity of demand as country Y.

The estimated premium being earned by the AWB(I) is then based on the simulated free market price, and the simulated average price for each class, as outlined in section 3.4. Estimating the latter involves identifying the aggregate value of the class of wheat, based on the simulated prices, and dividing by aggregate quantity. That is, the estimate of the premium is based on the systematic variation in prices that has been identified, and excludes any random elements that may have had a transitory impact on country specific prices.

The quantities used are the volumes traded by countries included in the regression analysis (i.e. those countries with greater than 10 contracts). Although this is a limitation, it should be noted that the original AWB(I) contract data only includes selected countries21.

3.5.3 DATA Using the data set previously discussed in section 3.4.3, a double log functional form is used. The log of the fob price ($US), is regressed against a series of monthly dummies (to abstract from the aggregate shifts in the market), and a series of country specific dummies using country Y as the baseline. The choice of country Y is purely for the reason that it has a high

21 These results, therefore, have been estimated on a different basis to those in the ACG report, which use all contracts.

69 volume of trades in all classes of wheat (its choice has no impact on the results). The equation also includes dummy variables for the grade of wheat traded, and the log of the exchange rate. The latter is specified as the local to $US exchange rate, and the coefficient is allowed to vary by country. The use of this bilateral exchange rate follows from that used by Carter et al (1999).

The exchange rate data used in this model is daily data for each country included in the analysis and was downloaded from FXHistory: Historical Currency Exchange Rates (http://www.oanda.com/convert/fxhistory) and is matched to each contract date. In identifying the preferred model, any insignificant exchange rate variables were dropped from the analysis. In some cases the exchange rate is relatively stable over the time period and this then suggests the variable may become collinear with the country specific dummy (that is the same variable is effectively being entered into the model twice). As a result, the standard error of the parameters on the country specific dummy variables and the exchange rate are inflated, which may lead to the conclusion that price discrimination does not exist when in fact it does.

Because of commercial in confidence requirements, only anonymised results can be reported, which are given in Appendix 2. Estimation results for the pricing to market study (Chapter 3), by classes of wheat (Class I, Class II etc.), are presented below. mon_1, mon_2 etc. are monthly dummy variables, where mon_1 equals 1 in January year 1, and zero otherwise: mon_2 =1 in February year 1, and so on. The country names have been replaced by (random) codes to maintain confidentiality. ERi denotes the exchange rate variable for country i. Country labels are consistent within equations (i.e. if a country specific dummy and exchange rate are both included, they can be identified as such). Only those that were significant at the 10% level were retained in the equation. Igrade_n are grade dummies.

22 3.5.4 RESULTS Table 3.4 reports the estimated premia per tonne for each class of wheat, over the 3 years. Note that all of these have been estimated using an estimated elasticity of demand of -8 for the baseline country, and that for each class of wheat this is taken to be the country with the lowest price. This is applied consistently across all three years of data. The only exception to this is for class 5 in 1997, where country E did not trade in that year. For 1997, country W is selected as the baseline country for class 5.

Table 3.4 Premiums received by class of wheat

Simulated price data, (US$/t)

Class 1997 1998 1999 Average 1 $1.61 $1.85 $2.03 $1.78 2 $1.36 $1.48 $1.55 $1.43 3 $0.11 $0.22 $0.33 $0.21 4 $0.59 $0.09 $0.01 $0.25 5 $0.16 $0.08 $0.56 $0.25 6 $0.79 $1.53 $2.14 $1.36 Average $1.00 $1.11 $1.09

It is important to note, that although not directly comparable due to different methods of simulation, note the differences between these results (table 3.4), and those reported by the AWB Ltd and ACG to the NCP Review. The AWB Ltd in house submission reported an average price premium of US$6.17 per tonne, and the ACG Report found an average maximum premium of US$1.38 per tonne. The estimates reported in this thesis are

22 Detailed results from the regression analysis are presented in Appendix 2

71 significantly smaller, due in part to the exclusion of price differentials which are here inferred to be a result of random variables as opposed to systematic price discrimination.

Table 3.5 reports the estimated aggregate value of the price premia, by multiplying the estimated premia by the quantities of each class traded:

Table 3.5 Total value of premium by class Based on countries with more than ten contracts traded, simulated price data (US$)

Class 1997 1998 1999 Total 1 $8,788,797 $10,440,666 $6,028,948 $25,258,411 2 $1,577,235 $1,010,881 $577,558 $3,165,674 3 $277,785 $663,041 $618,968 $1,559,794 4 $794,010 $133,421 $4,922 $932,353 5 $304,357 $72,052 $643,725 $1,020,134 6 $184,145 $336,225 $259,262 $779,632 Total $11,926,330 $12,656,286 $8,133,382 $32,715,998

It should be noted that these aggregate values are based only on the quantities traded to the countries included in the empirical model: those with more than ten contracts. Whilst it is tempting to simply extrapolate the value across all countries, this would be misleading, as it would assume that the small volume countries have a price distribution similar to the large volume countries. There is neither reason to assume that this is the case, nor that the average price premium could rise (or fall), as a result of including small volume countries. However, in cases where the volumes used are small, it is unlikely that their inclusion would greatly change the average premium or the aggregate value.

Further to the above results it may be of interest to compare these values with those that would be generated if the model was solved using the actual price differentials that is, to include in the solution price variability that is not systematically related to individual countries as if it had been achieved by purposeful actions by the AWB(I).

Table 3.6 reports the estimated premia per tonne for each class of wheat, over the 3 years.

Table 3.6 Premiums received by class

Actual price data, (US$/t)

Class 1997 1998 1999 Average 1 $2.02 $2.37 $1.95 $2.15 2 $1.43 $1.74 $1.65 $1.57 3 $0.39 $0.45 $0.41 $0.42 4 $0.54 $0.68 $0.50 $0.59 5 $0.18 $2.33 $0.64 $0.89 6 $0.39 $2.67 $4.52 $2.12 Average $1.25 $1.71 $1.20

In general, it can be seen that the price premia are larger when the actual price differentials are used to calibrate the equilibrium model as compared with using the simulated price differentials (which only include those statistically significant elements). Table 3.7 shows the results using the actual price data for the total value of the premium in US dollars (US$).

Table 3.7 Total value of premium

Based on countries with more than ten contracts traded,23 actual price data (US$)

Class 1997 1998 1999 Total 1 $11,043,198 $13,369,524 $5,789,465 $30,202,187 2 $1,667,226 $1,191,599 $613,454 $3,472,279 3 $952,152 $1,329,543 $763,124 $3,044,818 4 $728,964 $1,051,465 $446,816 $2,227,245 5 $329,057 $2,174,027 $730,867 $3,233,951 6 $89,465 $586,758 $547,588 $1,223,810 Total $14,810,061 $19,702,915 $8,891,314 $43,404,290

The graph (Figure 3.1) below shows the difference between the average price premiums received by class (1997-1999) using simulated and actual price data.

23 It should be noted that it is possible to decompose the aggregate change in premiums into the premium associated with each country, however, due to the confidential nature of the data, this information is commercially sensitive and cannot be reported.

Figure 3.1 Price premiums received by class

Based on actual price differences and simulated price differences

2.5 Simulated Actual

1.5

1 US$/tonne

0.5

0 1 2 3 4 5 6 Wheat Class

It can be seen that the premiums using the actual price data are much larger than for the simulated price data. This suggests that the price premiums have been generally over estimated due to the inclusion of non-systematic price effects with a positive bias.

3.5.5 CONCLUSION For each class and year the average premium, total value of the premium and predicted equilibrium price are reported. Re-solving the simulation model, using the devised Carter- Knetter framework, has lead to estimates of price premia which range from US$0.01 to US$2.14. For the contracts considered, the average value of the premiums is approximately

US$10million per year, half as large as the figure reported by ACG of US$21.5 million, and only a tenth of the AWB Ltd’s result of US$145 million. The average premium per tonne across all classes and years is US$1.02 per tonne, US$0.36 less than the ACG’s average premium per tonne. These results, reported in section 3.5.4, indicate that there are some statistically significant country specific effects for most classes of wheat traded which suggest some ability to price to market. These effects have manifested themselves either as country specific shifters, or as a significant relationship between the price being charged and the exchange rate of the country.

3.6 CONCLUSION The results from these analyses suggest that the functional form of the demand curve can have a large impact on the magnitude of the premiums. Using a linear demand curve increases the estimate of the premium generated by the simulation model, and for the data used here, this can be by a factor of 2-3 times. A priori there is no reason to assume that any specific functional form is correct, which raises some questions about the usefulness of the model if precise estimates of the premia are needed to evaluate the impact of the AWB(I). Alternative functional forms could reduce the estimate of the premia.

The pricing to market study and regression analysis indicates statistically significant country specific effects for most classes of wheat traded which suggest some ability to price to market. These effects manifested themselves either as country specific shifters, or a significant relationship between the price being charged and the exchange rate of the importing country.

Re-solving the simulation model leads to estimates of premia, which range up to US$2.14 per tonne. For the contracts considered, the average value of the premiums is approximately US$10 million per year. The average premium per tonne across all classes and years is US$ 1.02 per tonne.

As a consequence of the above analysis, it may be concluded that the AWB(I) has some ability to set prices for certain types of wheat in some overseas markets. This market power provides the rationale for the model of a price-setting monopolist developed in Chapter 5 and is used to investigate the impact on the, former, AWB, and, the now, AWB Ltd’s pricing behaviour following the change in firm objectives imposed by the Australian government’s drive for microeconomic reform.

CHAPTER 4 A POLITICAL ECONOMIC REVIEW OF MICROECONOMIC REFORM AND PRIVATISATION

4.1 INTRODUCTION The literature surrounding the behaviour of the firm, public enterprises, microeconomic reform and the privatisation process is both detailed and vast, making it difficult to provide a complete overview of the research undertaken in these areas. As a result, section 4.2 focuses on a political economic review of the process of microeconomic reform and privatisation, with specific reference to Australia. Section 4.3 concludes.

4.2 MICROECONOMIC REFORM AND THE PRIVATISATION PROCESS Ensuring efficiency and public welfare maximisation were deemed to be the traditional reasons for the existence of public enterprises. However, it is these motives for the existence of public enterprise that have ulitmately become the focus of their demise. This challenged the classical economist’s ‘laissez faire’ policies and, as a result, a transformation took place, and government owned and operated companies were seen as ‘inefficient’ and ‘obsolete’.

Microeconomic reform, and more specifically the privatisation process is an alternative to public enterprise, aimed at achieving the desired efficiency without government ownership or control.

Privatisation of public enterprises signals an advance of capitalistic thinking, as nationalization signaled an advance of socialistic thinking. It is the trust in the efficiency of markets, and the distrust in the efficiency of government…

(Bos, 1991, preface).

Privatisation, “the transfer of an enterprise from public to private ownership, either totally or partially” (Bos, 1986, p 31), is one of the main tools used to achieve reform with desired outcomes of increased internal and allocative efficiency.

Microeconomic reform and the process of privatisation has been a Western political phenomenon of the last three decades.

The breakdown of Keynesian economic policy during the 1970s, coupled with the 1979 recession, lead to a push towards microeconomic reform as a means to remove focus from the macroeconomic problems of the day

(Quiggin, 1996, p 11).

The US and the UK began to privatise key industries, such as telecommunications industries, during the late 1970s and early 1980s. The US reforms were ultimately limited to the airline, road and telecommunications industry, however the UK, under the conservative Thatcher government, began to embrace the concept of privatisation more widely (Bos, 1991, p 31 and Quiggin, 1996, p 12). At the same time, international financial deregulation was hastened by the collapse of the Bretton Woods agreement of fixed exchange rates.

The UK privatisation programme, initiated under the Thatcher government in the late 1970s, is one of the most extensive programmes in the world and many studies have been completed examining the processes undertaken to free up state ownership of utilities and other public sector enterprises. The aim of privatisation in the UK was to identify natural monopolies and to improve productive and allocative efficiency by provision of market forces (cost minimisation), induction of incentives for managers and workers, and increased competition for consumers (bringing consumer demand inline with marginal costs of supply) (Bishop et al, 1994, p 5). A comprehensive discussion of the UK privatisation programme

79 can be found in Vickers and Yarrow (1988) or for further information on British microeconomic reform see Bishop et al (1994). As with most leading academics in the field of public enterprises and privatisation Vickers and Yarrow (1988) and Bos (1986) focus on the standard public utilities, such as water, electricity, public transport, finance and education and health.

It is often suggested that Australia, as a result of abnormally high labour productivity and living standards developed in the 1880s, had ridden on the crest of a wave too long. In truth the Australian economy had been suffering from poor economic performance lagging behind most other western economies during the later part of the 20th century, “largely because of insufficient structural change” (Clark, 1995, p 145). Recognition of this came in the 1970s with the Whitlam government’s (1972-1975), 25% across-the-board tariff reduction in 1973, and the establishment of an Industries Assistance Commission (now the Productivity Commission).

By 1980 “the first systematic program of microeconomic reform” was presented in Australia at the Crossroads: Our Choices to the Year 2000 by Kasper, Blandy, Freebairn, Hocking and O’Neill (Quiggin, 1996, p 1). This paper presented a ‘libertarian’ and a ‘mercantilist’ path, which differed in their objectives and were supported by proponents for and against microeconomic reform (see table 4.2). Kasper et al., suggested the mercantilist approach to be the more direct option of microeconomic reform, however, it has been the libertarian path that has been chosen - a more conservative and long term option (Quiggin, 1996, p 2).

Table 4.1 Approaches to microeconomic reform in Australia

Libertarian Mercantilist Free international trade; Protection against import protection; Acceptance of structural changes wrought Protection against changes wrought by new by new technologies; technologies; Elimination of restrictions on international Maintenance of restrictions on capital capital flows and competitive domestic inflows and competitive capital markets; capital markets; Variation of wages in response to market Defence of a rigid system of occupational forces; and real wages; Reduction in the government’s benevolent Continuation of provisions by a benevolent role in service provision; government (e.g. health, education and welfare); Application of antimonopoly legislation Government by lobbying; and market deregulation; Expansion of the government’s role as a Consumerism and environmentalism provider of income maintenance. supported by bureaucratic regulation.

(Source: Kasper et al., 1980, pp 182-211, in Quiggin, 1996, pp 1-2).

The Liberal leader Fraser (1975-1983), is officially credited with instigating the ‘retreat of the State’ and his government is attributed as being instrumental in the future of Australian privatisation policy. Fraser’s economic policy goals were primarily macroeconomic in nature focusing on reducing inflation by bringing down real wage rates. Contrary to perception, these labour market reforms resulted in advancing the decline of the Australian government’s market intervention:

Although it was not recognised at the time, the abandonment of the full-employment objective had the effect of undermining government intervention in general. (Quiggin, 1996, p26).

On the microeconomic front specifically, Fraser established the Campbell Committee (1979) which initiated investigations into the Australian financial industry.

The Hawke-Keating Labour governments (1983-1996), against the grain of traditional Labour policies of state ownership, were also fundamental in Australia’s drive for deregulation and privatisation24. The intellectual (and bipartisan), environment in Australia during the 1980s fueled this change in Labour policy ideals (Quiggin, 1996, p28). This ‘new’ school of thought, encouraged by developments in the UK under Thatcher’s government, supported the principles of microeconomic reform with specific reference to tariff policy and financial deregulation (Quiggin, 1996, p28).

Several key events took place in the 1980s, cementing Australia’s stance in microeconomic policy procedure. In 1983, the Australian dollar was floated, initiating increased support for the deregulation process and continued microeconomic reform. The Campbell Committee proposals were implemented in 1986. Tariff reductions were formalized in 1988 and privatisation of key public holdings, such as the Commonwealth Bank, was instigated. By 1990 a competitive structure had been outlined for the airline and telecommunications industries (Quiggin, 1996, p 28). Australia had joined the international bandwagon of privatisation and deregulation.

24 However, there was a staunch difference between this Australian experience of privatisation and that of the UK or other economically developed nations. In Australia, microeconomic reform was perpetuated by a Labour government. The Australian Labour Party (ALP) had generally been a more left wing, socialist party interested in pursuing the rights of workers. This shift in policy, usually accounted for by “emphatically conservative governments” (Bos, 1986), has had interesting repercussions in Australian politics and economics.

Following the 1992 recession there was a lull in support for microeconomic reform, “the credibility of ‘economic rationalism’ was gravely reduced” and, “’reform fatigue’” had set in (Quiggin, 1996, p 29). Fraser (1991) proposed that perhaps lessons could be learnt from the UK privatisation experience (p 30). Firstly, Fraser suggested that judgment on a case by case basis should be considered as to whether privatisation is the best and most preferred policy, specifically in relation to the structure of the institution to be privatized and the degree of regulation that may be required in order for reform to be achieved (Fraser, 1991, p 30-32). Secondly, that transfer of ownership need be taken into consideration in Australia to ensure that projects were not financed by foreign debt, perhaps “open market purchase is the best option for Australia” (Fraser, 1991, p 35).

However, general skepticism was soon reversed with the launch of the Hilmer Report on competition policy (1993), aimed at ensuring that private enterprises were not being adversely affected by the lack of competition within the public sector (Quiggin, 1996, p 29). A direct consequence of the Hilmer report (1993), was the ratification of the National Competition Policy Act by the Council of Australian Governments (COAG) (1995). This Act saw the creation of the Australian Competition and Consumer Commission (ACCC) and the National Competition Council (NCC). The aim of the NCC is “to supervise the progress of federal and state governments towards implementation of competitive reform” (Quiggin, 1996, p 29). Since its establishment, the NCC has endorsed many industry reviews, including the 2000 Review of the Wheat Marketing Act.

Australia’s privatisation programme has been continued under Howard’s Liberal coalition government (1996 - to present). Howard’s aim has been to develop a policy agenda for “stronger sustainable growth, higher productivity, expanding opportunities and rising (sic) living standards” via a “major” privatisation platform focusing on competition, choice and efficiency (Howard, 1997). The Howard government’s policy has been sustained during the 1990s and has realised many of the Labour party’s legacies. The NCC’s National Competition Policy (NCP) Review programme has investigated many industries across all

83 sectors, including transport, health, retail, education, primary industries, water reform, legal, financial services and business (See www.ncc.gov.au).

The late 1990s saw substantial moves towards privatising more key industries within Australia. Action to privatise the telecommunications industry, a controversial issue since Keating’s government, is now underway (2003) (www.aph.gov.au), as is the privatisation of the Totalisator Agency Board, various state based public transport operations (e.g. rail in Victoria), and the postal system.

The pace of microeconomic reform looks set to continue as an integral component of Australian federal and state government policy, even as the, often negative, impacts of such programmes are becoming increasingly apparent in other western nations.

4.3 CONCLUSION As a result of the western political trend for microeconomic reform, the process in Australia has extended across a wide variety of industries including agriculture. Privatization of the wheat industry’s largest and most powerful player, the AWB Ltd, discussed in detail in Chapter 2, has provided a basis for the investigation of the effects of ownership on firm behaviour. Application of the shift of the AWB from a government owned and operated public enterprise to private firm allows for a novel examination of a multi-market firm’s pricing policies under different ownership structures. This analysis is undertaken in the next chapter.

CHAPTER 5 MODELLING THE BEHAVIOUR OF THE AUSTRALIAN WHEAT BOARD

5.1 INTRODUCTION The model developed in this chapter is based on that outlined in Fraser (1989) of a size- orientated price-setting firm operating in multiple markets. Fraser (1989) examined the public enterprise’s pricing policy given risk averse management and the production of more than one output (Fraser, 1989, p 148). In that case little was done to specify the firm’s alternative markets other than for them to differ in terms of the extent to which demand was uncertain.

Bos (1986) notes that the study of public enterprise economics, focusing specifically on pricing, is indeed different to the study of pricing in private firms and this difference is not ownership, “The main difference is the multitude of political and economic determinants of public enterprises’ activities as compared to the mainly commercial determinants of the activities of private enterprises”…”Prices are the best indicator of the consequences of combining such political and economic determinants” (Bos, 1986, p 13-14). From this it can be acknowledged that the transformation from public to private ownership may have a substantial impact on both the pricing behaviour of a firm but also on its political and economic influences.

Fraser’s (1989) model, extends the work of Rees (1984). Rees (1984) models a public enterprise being directed by risk neutral managers whose objective function is to choose price to maximise output under uncertain demand and production. Rees’ (1984) analysis is extended “by demonstrating the importance of not only the expected profit constraint but also the attitude to risk of managers in determining a public enterprise’s relative prices” (Fraser, 1989, p 149).

Section 5.2 outlines and qualifies the model’s assumptions. Section 5.3 provides a more detailed analysis of the demand functions used in the modelling process and section 5.4 formally develops the model. Preliminary numerical results are presented in section 5.5 and 5.6 concludes.

5.2 ASSUMPTIONS OF THE MODEL In this case, with a price setting firm operating in multiple markets, it is important to characterise more fully than in Fraser (1989) the differences between the firm’s domestic and overseas markets. The following market-based assumptions are made:

(a) The product is a homogenous good; (b) three markets exist: ‘overseas’, ‘domestic’ and residual production (‘sump market’) in which revenue just covers costs; (c) costs to supply are greater in the overseas market than the domestic; (d) demand in the overseas market is more elastic than in the domestic market; (e) demand in the overseas market is more uncertain than in the domestic market.

Other important assumptions include the AWB Ltd’s ability to price set and that they operate as a risk averse firm. These assumptions are more contentious than the market based assumptions listed above.

5.2.1 WHEAT AS A HOMOGENOUS GOOD The question of ‘Is wheat just wheat?’, has been avidly pursued by agricultural economists over the last twenty years. Recent literature sees wheat investigated in both frameworks, either as a single product or differentiated by quality or variety (Larue and Lapan, 1992; Ahamdi-Esfahani and Stanmore, 1992; Wilson, 1994). Although economists often view wheat as a differentiated product, government bodies, private companies and other data collection agencies tend to refer to wheat as an aggregate commodity, whether this be for

86 ease of collection, or because of the commercial sensitivity of disaggregate data25. These data restrictions are the main reason for the use of wheat as a homogenous commodity in economic studies, and this thesis is no exception.

Generally however, it is agreed that realistically and scientifically wheat is more than simply a homogenous product and hence it deserves to be analysed as a heterogeneous good by class or grade as opposed to in aggregate. The principal economic reason for this is because buyers import different classes of wheat in order that the attributes of the wheat match with those required for diverse end uses. Not only are different qualities of wheat demanded, but farmers aim to produce wheats that suit the physical constraints of their land or climate. Sellers and marketers are then able to take these different wheat varieties and classify and segregate the wheat by inherent or physical attributes. In turn these classes can signal quality to buyers and as a result sellers can demand premiums depending on the quality characteristics of their wheat.

5.2.2 THREE MARKETS This model is assumed to have three markets, an overseas (or export) market, a domestic market and a ‘sump’ market. This assumption fits well with the structure of the AWB Ltd, who are the sole marketer and seller of Australian wheat for export and a principal seller on the Australian domestic market. The residual market has been introduced as three markets are a more realistic assumption and this structure provides us with ease in the modelling process, as the firm is not solely constrained to selling to two markets.

5.2.3 COSTS Costs of selling wheat internationally will be higher due to the marketing effort required to secure sale and the costs of transporting wheat to its destination. More wheat is being sold

25 AWB Ltd only allows data to be reported in aggregate form in order to protect their commercial operations.

‘cost insurance and freight’ (cif) than previously, which requires the AWB Ltd to foot all costs associated with shipping. For a more detailed analysis on costs, see chapter 5.

5.2.4 ELASTICITY OF DEMAND The elasticity of demand for Australian wheat has not been directly examined either on the domestic or export market primarily due to data restrictions. Large scale general equilibrium models are often used to attempt to calculate elasticities of demand for many products, across a wide range of markets. Price elasticity of demand for Australian wheat ha ve been quoted at –8, based on estimates from the MONASH model, the Murphy general equilibrium model and GTAP (ACG Appendix B, 2000, p 75-76). A study by ERS/Penn State/WTO (Stout and Alber, 2003) suggest, using a large scale, non-spatial, comparative statics model, that own price elasticity of demand for Australian wheat domestically is -0.034, which is confirmed by the OECD (Stout and Alber, 2003, Table 19).

Some indirect measures of elasticity of demand for Australian wheat have also been conducted such as Carter et al. (1999) and Saris and Freebairn (1983). These papers report figures of -0.08 using residual inverse demand elasticities for Australian wheat exported to Japan, and -0.10 for elasticity of demand for domestic wheat, respectively.

After brief examination this cross section of studies on the elasticity of demand estimates for Australian wheat, it can be assumed that the results are generally inconclusive. Following from economic theory, it is widely accepted that, land confirmed by the above analyses, agricultural goods are more inelastic than non-agricultural goods and that the overseas market is more elastic than their smaller domestic counterparts. This assumption is especially applicable to Australia as the Australian domestic wheat market deals with approximately two thirds less total volume than Australian wheat destined for the export market. The parameter values used in this thesis have been arbitrarily set at -1.3 for the domestic market and -1.5 for the export market. It was important to keep the two figures similar due to model specification.

26 5.2.5 UNCERTAINTY OF DEMAND Demand is assumed to be more uncertain in the overseas market than in the domestic market. This is generally true, as the AWB Ltd will have much less available information on their overseas customers and their demand for wheat than will be available in the domestic market. Volatility in the international market is also more prevalent and changes in tastes and preferences, an important condition of demand, are likely to be more influential in the export market. Competition and availability of substitute goods will also have an affect on the level of demand uncertainty; again this is likely to be more prevalent in the international market. For further discussion on uncertainty see chapter 6.

27 5.2.6 AWB LTD AS A PRICE SETTING FIRM The assumption that the AWB Ltd behaves as a price setter is the most difficult assumption to qualify. Prior to 1989, the AWB was operating as a statutory marketing authority under federal government regulation. Behaving as a revenue maximising public enterprise, AWB had full market power on the domestic market. Since deregulation of the domestic market there have been very few studies examining this issue, although the general opinion is that AWB Ltd still holds a substantial share of trade in the domestic market (see Wait and Ahmadi-Esfahani, 1996).

Several studies on the AWB’s market power in the international arena have been examined and the consensus is very much inconclusive (see chapter 2). Essentially data restrictions are the major reason for the lack of economic studies in this area, and also help to explain the wide spectrum of results. The majority of the studies investigated are qualitative in nature and use economic theory to determine Australia’s position in the structure of the world wheat market. However, key to justifying this assumption, results from the empirical

26 See Horowitz, I. (1970), Decision Making and the Theory of the Firm, Holt, Rinehart and Winston, Inc, USA, pp 76-93. 27 For further details on previous literature see chapters 2 and 3.

89 work (based on the Carter – Knetter Model) presented in chapter 3 suggests that there is some evidence of price premia being earnt by the AWB(I).

It is interesting to note, that the AWB Ltd feels, with respect to retaining single desk arrangements, that they command significant market power. Economic theory suggests the contrary that as Australia is only the fourth largest exporter of wheat with approximately 12% total market share, and they are not likely to have an ability to price discriminate on the international market. However, until publicly available studies are conducted to confirm or deny the existence of market power, it is plausible, following from the results presented in Chapter 3, that the AWB Ltd has sufficient ability to set prices in overseas markets.

From a basic numerical analysis the AWB Ltd has, on average, since 1992/93, commanded approximately 50% of domestic market share and hence has the capacity to price discriminate in the Australian domestic market28.

5.2.7 AWB LTD AS A RISK AVERSE FIRM The AWB is assumed to be risk averse because of its dependence on growers, as principle shareholders, and due to the use of futures markets to hedge the risk faced in the international wheat market. Farmers are often seen as risk averse players due to their inability to control neither their production of a good nor their return. This in itself helps to justify why the majority of farmers (85 per cent of grain growers, including 86 per cent of major accounts29), are in support of the single desk in Australia. The opinion is that the single desk provides them with some assurance as to the returns they will receive for their wheat and the other benefits, such as market information, that would otherwise not be provided.

28 Using data from ABARE, 2001 (Australian domestic use (p 216)), and www.awb.com.au (AWB receivals and net exports). Also, quoted in Kronos Corporate (2002): “AWB claim 50% domestic market share (A. Lindberg, FMCA Conference, 2002)”, (Footnote 30, p 30). 29 AWB’s research cited in Australian Financial Review, 17th January, 2001

AWB(I)’s reliance on the Sydney Futures Market as well as the Chicago Futures Market illustrates that AWB(I) has a reason to insure themselves against price and market risk (e.g. exchange rates).

AWB’s risk management strategy involves determining the level of grain price, foreign exchange and interest rate exposure and establishing a policy setting to reduce exposures to acceptable levels. The objective of this policy is to protect the financial viability of the wheat pools through application of prudent hedge cover. The overriding risk management objective is to maximise favourable opportunities to enhance pool returns.

As the sole exporter of Australian wheat, AWB is responsible for managing the price risk on 10 to 20 million tonnes of wheat annually. This makes AWB one of the largest managers of commodity price risk in the world. In 1998/99 AWB’s commodity futures hedging position totalled four million tonnes, the largest position it has recorded. While virtually all wheat is received by AWB at harvest, it is sold over the course of the year, and the price risk is managed from well before harvest and on through the season.

AWB uses numerous techniques to manage price risk, using both physical and derivative wheat markets. While AWB is one of the largest users of US agricultural futures and options, the most important method of managing price risk and securing pool returns for growers is through the physical sales program (sic).

(AWB Ltd, 1999)

5.3 THE MODEL Regarding the specification of the firm’s objective, based on Fraser (1989), it is assumed that “before” commercialisation the objective of the firm is to maximize the expected utility of sales revenue (EU(Rev)T) subject to an expected profit constraint (E(P )T) and a total production constraint ( Q ). Note that in what follows consideration of revenue from the sale

91 of residual production is omitted in order to simplify the analysis. In this context it can be shown that because the firm’s pricing behaviour is always constrained, this residual revenue source has a negligible effect on behaviour, even if the firm is very risk averse. In this case the firm’s objective is given by:

Max EU (Re v) …(5.1)

By choice of overseas (po) and domestic (pd) prices. Subject to:

Q = E (qo ) + E(qd ) + qx …(5.2) and,

E(Õ)T ³ z …(5.3)

E(P)T = po E(qo ) - co E(qo ) + pd E(qd ) - cd E (qd ) …(5.4) where:

E(qo) = expected sales in the overseas market;

E(qd) = expected sales in the domestic market; qx = sales of residual production; co = costs of supply per unit to the overseas market; cd = costs of supply per unit to the domestic market; z = minimum feasible expected profit level.

Demand in both the overseas and domestic markets is assumed to be characterised by 30 constant elasticity (bi) demand functions subject to additive uncertainty (ui) .

30 See section 5.4.2 for more details regarding the demand function specification.

-bi qi = ai pi + ui where i=o or d …(5.5)

Price is chosen as an optimal mark-up (l ) on per unit costs of supply:

pi = (1+li)ci …(5.6) where:

ai = scaling factor in each market

E(ui ) = 0 and demand is assumed to be uncorrelated in the two markets. As a consequence:

-bi E(qi ) = ai pi …(5.7)

and expected revenue (E(Rev)T) is given by:

E(Re v)T = poE(qo ) + pd E(qd ) …(5.8)

While the variance of revenue (Var(Rev)T) is given by:

2 2 Var(Re v)T = po Var(uo ) + pd Var(ud ) …(5.9) where:

Var(ui) = variance of demand in each market.

On this basis, using the mean-variance specification of expected utility, the firm’s objective is to maximize31:

31 See Hanson and Ladd, 1991 for empirical support for this assumption.

1 EU (Rev) = U(E(Re v) ) + U ''(E(Re v) )Var(Rev) …(5.10) T T 2 T T Subject to:

Q = E (qo ) + E(qd ) + qx …(5.11) and,

E(Õ)T ³ z

E(Õ)T = ( po E( qo ) - coE(qo )) + ( pd E(qd ) - cd E(qd ) …(5.12)

The first order cond itions for the optimal prices, subject to the expected profit and total production constraints are as follows:

1 1 foc = U ¢(E (Rev)) * E¢(Rev) + [U ¢¢¢(E(Re v)) *Var Re v * E¢(Re v) ] + [U ¢¢(E(Re v)) *Var¢ Re v ] = 0 i T i 2 T T i 2 T i

…(5.13) where,

E¢(Rev)i = ci (E(qi ))(1-bi )

…(5.14)

Var¢Revi = ci pi (Var(ui )) …(5.15)

The “after” commercialization situation is assumed to be represented by a focus on profit rather than revenue, in which case the firm’s objective is to maximize the expected utility of profit, subject only to the total production constraint (Fraser, 1994(a)).

Max EU (P)T = E(U (Spi E(qi ) - Sci E(qi ))) …(5.16)

Subject to:

Q = E(qo ) + E(qd ) + qx …(5.17)

Using the same specification of the demand functions, expected profit is given by:

E(Õ)T = po E( qo ) + pd E(qd ) - co E(qo) - cd E(qd ) …(5.18)

and the variance of profit is given by:

2 2 Var(Õ)T =(po -co) Var(uo) + (pd - cd) Var(ud) …(5.19)

Once again using the mean-variance formulation gives:

1 Max EU (P) = U (E(P) + U ¢¢(E (P) )Var(P) …(5.20) T T 2 T T Subject to:

Q = E (qo ) + E(qd ) + qx …(5.21)

On this basis, the first order conditions for the optimal prices subject to the total production constraint are given by:

1 1 foc =U¢(E(P)) *E¢(P) + [U¢¢¢(E(P)) *VarP *E¢(P) ]+ [U¢¢(E(P)) *Var¢P ] = 0 i T i 2 T T i 2 T i

…(5.22) where,

2 -bi -1 E¢(P)i = ciE(qi ) -libici ai ((1+li )ci ) …(5.23)

2 Var¢Pi = 2lici Var(ui ) …(5.24)

This completes the specification of the model on which the following preliminary numerical analysis is based.

5.4 NUMERICAL ANALYSIS In order to undertake a numerical analysis of the model developed in the previous section is it necessary to specify a functional form for the firm’s utility function, and a set of base case parameter values. In what follows, use is made of the constant relative risk aversion utility function (see Fraser 1994a and 1994b). On this basis, total utility for the “before” commercialization case (U(Re v)T ) is given by:

(Re v) 1-R U(Re v) = T …(5.25) T 1- R

And in the “after” commercialization case, the firm’s utility is given by:

1-R PT U (P)T = 1 - R …(5.26)

The parameter values used for the ‘base case’ in the numerical analysis are as follows:

Overseas Market Domestic Market Residual Market ao = 10000 ad = 10000 px = cx = 1 bo = 1.5 bd = 1.3 Q = 240 co = 10 cd = 10 uo = 5 ud = 5

Note that the only difference in the characterization of the two markets in the base case is in the elasticity of demand, with demand being more elastic in the overseas market. In addition, the relative risk aversion coefficient is set at R = 0.5 and the expected profit constraint (z) for the expected utility of revenue maximiser is set at 95% of that achieved by the expected utility of profit maximiser. This is an arbitrary assumption, which is made for simplicity, and in order to keep the two types of pricing behaviour “close” to each other.

On this basis, Table 5.1 shows the “before” and “after” scenarios for the firm.

Table 5.1 Simulated results for prices and quantities as a result of a change in objectives

Differences in elasticities

Before After

bo=1.5 bo=1.5

bd=1.3 bd=1.3

po $24.00 $30.00

pd $26.30 $43.30

QT 227.6 135.4

E(P)Total $3514.77 $3700.35

This table shows that the shift to a profit-orientated objective results in an increase in price in both markets, with an associated decrease in sales overall. Note that the increase in price is greater in the less elastic market32.

Table 5.2, shows the impact of allowing for differences in the costs of supply on pricing behaviour. Elasticities and variances are held at the same levels as in the first case.

Table 5.2 Simulated results for prices and quantities as a result of a change in objectives

Differences in costs and elasticities

Before After

co=15 co=15

cd=10 cd=10

po $32.70 $44.85

pd $27.60 $43.30

QT 187.4 107.9

E(P)Total $3303.41 $3476.99

The results for case 2 differ to those in the first case with po>pd for both objectives. However, there are also similar movements in prices and sales between the before and after scenarios.

32 Note also that the expected utility of revenue maximiser will choose to lower prices until it is constrained by the expected profit constraint. Because of this the first order conditions are not equal to zero for the expected utility of revenue maximiser. However, they must be equal to each other in order for the best contribution to be made to increasing the expect utility of the revenue maximiser.

Table 5.3, indicates the impacts that occur where elasticities and costs differ between the two markets as with the previous scenario (bo=1.5, bd=1.3 and co=15, cd=10) but with the inclusion of the final market difference: a difference in the variances on the demand functions, with this variance being greater for the overseas market (Var(uo)=500, Var(ud)=5).

Table 5.3 Simulated results for prices and quantities as a result of a change in objectives

Differences in costs, elasticities and variances

Before After

Var(uo)=500 Var(uo)=500

Var(ud)=5 Var(ud)=5

po $33.00 $41.10

pd $27.40 $43.30

QT 187.9 112.5

E(P)Total $3301.71 $3473.74

With this increase in the variance of the overseas market’s demand there is a minimal impact on the pricing behaviour of the expected utility of revenue maximiser compared with the results for the prices in case 2. This is due to the fact that this firm’s choice of prices is constrained by the expected profit constraint. Whereas, in the “after” scenario, the firm is free to adjust its prices to reflect the increased demand uncertainty in the overseas market. Given the firm’s risk aversion, this results in the overseas market being perceived as less attractive, and the price set for that market is lowered in order to reduce the variability of

99 profits. Note that this decrease is sufficient to reverse the relative level of domestic and overseas prices for the expected utility of profit maximiser from that in case 2.

5.4.1 SENSITIVITY ANALYSIS In order to examine these issues further a sensitivity analysis of the impacts of different levels of risk aversion was undertaken. The results of such an analysis are recorded below in Table 5.4.

Table 5.4 Sensitivity to changes in the relative risk aversion coefficient

Before After R=0.1

po $32.70 $43.90

pd $27.60 $43.30

QT 187.4 108.9 R=0.9

po $33.20 $39.10

pd $27.10 $43.30

QT 189.7 115.4

The results in Table 5.4 show once again the insensitivity of the pricing behaviour of the expected utility of revenue maximiser, although there is a slight inclination for the more risk averse firm to concentrate on increasing sales (by lowering price), in the less risky (domestic) market. The effects for the expected utility of profit maximiser indicate that varying risk aversion reverses the rankings of po and pd which is consistent with the sensitivity of this ratio to the variance of demand identified in Table 5.3. Nevertheless, the

100 previous findings that the shift from a size-orientated objective to a profit-orientated objective results in increases in prices in both markets, and an associated decrease in total sales, remains robust.

5.4.2 DEMAND FUNCTIONS The model was originally developed with linear demand functions however, initial analysis (with no uncertainty coefficients), suggested that the elasticities of demand were not having any impact on the results, and the model was being driven solely by costs. Upon making this discovery the demand curves were switched to a constant elasticity demand functions (equation (5.1)) as the linear demand curves were not allowing for changes in elasticities to be represented.

Following this decision, it was important to determine an additive or a multiplicative structure regarding the intercept term (ai).

Equation (5.1) was implemented assuming that elasticity in the overseas market was greater than in the domestic market, and that pi = (1 + li )ci .

After analysis of preliminary results using the linear and the additive constant elasticity demand function, it was discovered that either costs (ci), or elasticities (bi), impact is dependent on constant elasticity or linear demand curves respectively. There appeared to be no obvious explanation for this behaviour.

A constant elasticity demand framework with a multiplicative intercept term was then investigated (this form of constant elasticity demand is generally perceived as more realistic).

101

Introduction of uncertainty in the demand function is a key component of the model used in this thesis. In order to transform the demand functions to include uncertainty, each type of framework (linear and additive or multiplicative constant elasticity) needed to be examined

Following from this investigation, a constant elasticity demand function was defined being multiplicative in the intercept term (ai) and either an additive or multiplicative uncertainty coefficient. The final demand function used in the model, as given in equation (5.1), uses an additive uncertainty term.

This form of constant elasticity demand model is more complex and possibly a more ‘real’ representation, and also allows for the function to be linearised (by taking logs) for empirical studies. However, with an additive component, the demand function will be intrinsically non-linear in parameters in econometric terms (Gujarati, 2002, p 177). Another benefit of using additive uncertainty (equation (5.1)), is that in taking expectations of demand, the expectation of an additive uncertainty coefficient (E(ui)), equates to zero, providing ease of calculation of the first order conditions. It was determined that the benefit of additive uncertainty was deemed greater than the benefit of a multiplicative uncertainty term.

It should be noted that in Chapter 3, the ACG’s report for the NCP Review process was examined and some analysis on the importance of the functional form of demand curve was conducted. It was determined that the use of the linear demand equation consistently leads to higher simulated premia, and that the function of the demand curve does indeed matter. Regardless of the results of this analysis, constant elasticity demand curves remain the preferred option for this model due to their simplicity and convenience. It should be suggested that interpretation of the results should be made with this possible limitation in mind.

102

5.4.3 HYPOTHESES It follows from the numerical analysis that the following three hypotheses can be developed:

H1: po (P) > po (Re v)

H2: pd (P) > pd (Re v)

H3: QT (P) < QT (Re v)

Based on the numerical analysis, there are also several ambiguities regarding relative price levels, whereby relative prices were shown to be dependent on differences in elasticity and in costs, and in the variance of demand or the risk aversion of the firm. In particular:

po (Rev) >or < pd(Re v)

In the numerical analysis, this ratio was shown to be dependent on market-based differences elasticities and costs of supply.

In addition:

po (Õ) > or < pd (Õ)

The above ratio was also shown to be ambiguous and dependent on market-based differences between elasticities, costs and on both the variance of demand in each market and the risk aversion of the firm.

5.5 CONCLUSION A model has been developed to investigate the impact on multi-market pricing behaviour as the objective of a firm is shifted from a revenue-orientated public enterprise to a semi-

103 regulated profit maximiser. Empirically testable hypotheses have been developed through the algebraic and numerical analysis of a risk averse firm’s price-setting behaviour for two different objective functions and given differing costs of supply, uncertainty of demand and differing price elasticities of demand for the firm’s markets.

Adaptation of Fraser’s (1989) modelling framework was outlined in 5.4, following a discussion of the assumptions of the model in 5.2. The model is of a size-orientated price- setting firm operating in multiple and segmented markets. These markets are specified to capture the differences between the AWB Ltd’s overseas and domestic markets. The overseas market is characterised as being a higher cost market, with more elastic and more uncertain demand than the domestic market. With this structure, the model incorporates a “before” and “after” commercialisation pair of objectives for the monopolist, where revenue and profits are the two objectives respectively.

Results of the numerical analysis were presented in 5.5. Three empirically-testable hypotheses were generated identifying the likely impact on the AWB Ltd’s overseas and domestic pricing behaviour of the changes imposed on it by the Australian government. In particular, it was suggested that the impact of commercialisation would have been to increase prices in both domestic and overseas markets, with an associated decrease in total sales. This section also showed how the change in objective affects optimal prices when the firm’s markets differ in each respect as well as the combined effect of all differences. This and a further sensitivity analysis of the effect of the firm’s level of risk aversion was conducted which confirmed the robustness of the three hypotheses, but also indicated a set of inconclusive results that will require further attention in an empirical context.

Chapter 6 will further develop this model applying the changes in firm objectives to concurrent industry related changes such as domestic market deregulation, transport cost developments and levels of international uncertainty which may simultaneously impact on the pricing behaviour of the AWB Ltd.

104

CHAPTER 6 IMPLICATIONS OF RECENT INDUSTRY DEVELOPMENTS FOR DOMESTIC AND OVERSEAS PRICES

6.1 INTRODUCTION The Australian wheat industry has undergone many changes over the last decade (1990s) including deregulation of the Australian wheat market; changes in domestic and export transport costs; and developments in international stability. These changes are imposed on the hypothetical data used in the original model to determine how pricing behaviour differs from the base case. The results for each scenario are then analysed and conclusions drawn. Sections 6.2, 6.3 and 6.4 outlines each recent Australian wheat industry development and analyses the results of these impacts on the pricing behaviour of the AWB as it shifts from a revenue maximiser to a profit maximiser. Section 6.5 provides a sensitivity analysis of the robustness of the hypotheses and 6.6 concludes the discussion.

6.2 DOMESTIC DEREGULATION OF THE AUSTRALIAN WHEAT MARKET The Australian Wheat Board (AWB), was originally established as a temporary measure during WWII, “to handle wheat marketing as a war-time emergency” (AWB, 1999). These terms were formalised with the introduction of the Wheat Industry Stabilisation Act (WISA), (1948). The act ensured the former AWB was the sole marketer and seller of Australian wheat on the domestic and export markets. As described by Wait and Ahmadi-Esfahani (1996), prior to deregulation:

105

All domestically produced wheat became the property of the AWB once it left the farm gate. The Wheat (sic) was then taken to the AWB-appointed receiver in each State – the Bulk Handling Authorities (BHAs), which were statutory monopolies. Growers were charged for the use of the services of the BHAs at the same amount per tonne regardless of the handling facility to which they delivered their wheat and the time of delivery within the season.

(Wait and Ahamadi-Esfahani, 1996, p 318).

This behaviour is known as cost pooling and was adopted in an attempt to ensure that farmers were not exploited by the rise of private sector monopolies (Wait and Ahamadi- Esfahani, 1996, p 318). Although producers were being ‘protected’ by WISA (1948), costs to consumers were high as the AWB was able to sell all wheat for domestic (human or animal consumption), at the same price regardless of its quality or end use.

During the late 1970s and 1980s several investigations by the former Industry Assistance Commission (IAC), took place challenging the legal arrangements of wheat marketing in Australia33. The concerns addressed included increases in marketing costs (specifically handling, storage, and transport costs), lower wheat prices, disadvantages for interrelated sectors, (such as the domestic livestock industry), and the possibility of complete deregulation of the wheat industry.

The preliminaries of deregulation followed the IAC’s reports and the announcement by the Australian Federal Government (1985), that they were no longer willing or able to provide assistance to wheat producers in order to match the subsidies provided to farmers in other countries (www.prairecentre.org/wheataustralia.htm). The Wheat Marketing Act (WMA), (1984, 1989), was developed to replace WISA (1948). Changes included the removal of the provision of government underwriting of loans and price guarantees for the AWB as well as opening the domestic market to competition to increase internal and allocative efficiency.

33 See IAC Reports, 1977, 1978, 1984, 1988(a) and 1988(b).

106

The deregulation process culminated with the signing of the WMA on 1st July 1989. The WMA imposed a new structure for the Australian wheat industry and the AWB became merely one of several players in the newly competitive domestic market. Multinational companies took this opportunity to enter the Australian market. The companies that began marketing, trading and broking in the domestic market included Cargill, Conagara and Louis Dreyfus (Wait and Ahmadi-Esfahani, 1996, p 320). Farmers were no longer restricted to selling solely to the AWB and now had marketing choices for domestic sales. Buyers also benefited from the increased competition in the marketplace (Wait and Ahamadi-Esfahani, 1996, p 320). For a more detailed analysis of the deregulation process see chapters 2 and 4.

On a theoretical level, domestic market deregulation may ha ve resulted in an increase in the elasticity of the AWB’s domestic demand as consumers would not be as constrained to purchasing wheat from the AWB as they had been prior to deregulation. It is important to note that although this is a widely accepted theoretical construct there have been no studies examining the elasticity of demand for the domestic wheat market as the data required is deemed commercially sensitive and has not to date been released by marketing agents (Wait and Ahmadi-Esfahani, 1996, p 321).

For this study, the increase in the elasticity of domestic demand is represented by a 0.1 unit 34 increase in bd to 1.4, from 1.3 in the base case . Costs and uncertainty are held the same as in the base case. The elasticity of demand for the international market also remains unchanged.

34 It is important to note that the change in elasticity of demand may be either larger or smaller, and that this change of 0.1 unit is arbitrarily used to investigate the impact of a change in elasticity of demand on the hypothetical data. However, unreported numerical simulations suggest that this merely affects the magnitude and not the direction of the results presented below.

107

6.2.1 RESULTS

Table 6.1 Comparing sales maximisation results when bd is increased

Before* Before

bo=1.5 bo=1.5

bd=1.3 bd=1.4

po $32.90 $30.87

pd $27.30 $24.73

QT 188.83 170.37

E(P)Total $3298.32 $2576.06

Where ‘Before*’ indicates the scenario results from the full model version of the base case (as in chapter 5, table 5.1).

As expected, the results in Table 6.1 indicate that an increase in the AWB’s elasticity of domestic demand, by weakening the expected profit constraint, will lead to a decline in both 35 its overseas and domestic prices . Note also that the domestic price has decreased by a greater amount than the overseas price as the revenue maximiser concentrates on using the weaker expected profit constraint to pursue increased sales revenue in the domestic market.

35 Note that the expected utility of revenue maximiser will choose to lower prices until it is constrained by the expected profit constraint. Because of this the first order conditions (focs), are not equal to zero for the expected utility of revenue maximiser. However, the focs must equate in order for the best contribution to be made to increasing the expect utility of the revenue maximiser.

108

Table 6.2 Comparing profit maximisation results when bd is increased

After* After

bo=1.5 bo=1.5

bd=1.3 bd=1.4

po $41.03 $40.32

pd $42.49 $34.50

QT 114.48 109.37

E(P)Total $3473.42 $2711.71

Table 6.2 again shows a decline in prices from the base case scenario. The decrease in the domestic price is considerably larger than the decline in the overseas price and this decline is significantly greater than in Table 6.1 above. This reflects the greater price flexibility associated with the (unconstrained) profit maximiser. Nevertheless, the results are consistent with the previous findings that overseas and domestic prices of the profit maximiser (after scenario) are still greater than for the revenue maximiser (before scenario), and that the quantity for the profit maximiser is less than for the revenue maximiser (see hypotheses, chapter 5).

These results suggest that if the effects of deregulation in the domestic wheat market appeared in advance of the commercialisation of the AWB being implemented then domestic consumers would have seen this in terms of a decrease in domestic prices until the implementation of commercialisation brought about a price increase. Alternatively, if the impact of deregulation appeared in conjunction with the impact of commercialisation, then no such price cut would be observed. Rather, the extent of the increase in the domestic price associated with commercialisation would simply have been reduced.

109

6.3 CHANGES IN TRANSPORT COSTS The Australian transportation industry (road, rail, and sea), and bulk handling corporations are closely linked with the wheat industry as they provide an integral network between farmers and consumers. Transportation has remained a key cost component post deregulation of the AWB as the majority of wheat (85%) is destined for the export market and requires transportation from receival points to port facilities. Domestic wheat sales account for the remaining 15% where the principal modes of transport are rail and road.

Rail is typically used for the transportation of non-perishable bulk commodities such as minerals or agricultural goods with Australian railways hauling up to 70% of all domestic grain including 80% of Australian export wheat (ARA, 200236). Rail is the second largest mode of freight transportation in Australia with 26% of total freight hauled (Austroads, 2000, p 16). Road transportation is the largest component of domestic freight transport in Australia, 65% of all freight moved in 1995/96 was by road and this figure has been increasing at 6% per annum over the last few years (Austroads, 2000, p 15 -16). Figure 6.1 below shows the increases in domestic freight transportation over the 1980s and 1990s as a billion tonnes moved per kilometre by mode.

36 Australasian Railway Association Inc (2002), “From Farm Gate to Plate”, Rail Fact Sheet #18, www.ara.net.au

110

Figure 6.1 Domestic freight transport

(1980/81-1994/95)

140

120

100

80 Road Government Rail 60 Private Rail 40

20 Billion tonnes per kilometre

1980-81 1982-83 1984-85 1986-87 1988-89 1990-91 1992-93 1994-95 Years

Source: Austroads, 2000, p 15 (Table 1.10).

6.3.1 DOMESTIC COSTS There are two main stages of grain transportation. Firstly, grain is transported from the farm to storage facilities (on average 17 kilometres), this cost is usually borne by the farmer (AWB Ltd, 2001a). Secondly, wheat is transported from receival storage facilities to domestic customers or merchants (a national average of 350 kilometres), the costs of which are usually incurred by the marketer, AWB Ltd (AWB Ltd, 2001a).

111

State governments have tended to approach transport matters parochially. As noted by the Australian Wheat Board and the Australian Shipping Users Group, the existing pattern of transport infrastructure largely reflects bias as the location of ports, bulk loading facilities, and rail lines has been largely determined on a state-by-state basis. The physical location of transport facilities has been supported by regulations relating to the road and rail shipment of cargo. These have tended to favour rail over road, and to prevent interstate shipment of certain commodities.

(IC, 1993, p 77).

RAIL AND ROAD The main form of transport for wheat from receival point to export terminal is rail. The Australian rail network is relatively extensive and the expansion of the freight and passenger rail systems has been inexorably linked with the Australian wheat industry. It is this historical link, the bulk characteristics of wheat, and the location of production areas that ensures rail has a natural competitive advantage over road transportation in the Eastern states (AWB Ltd, 2001a).

The rail industry in Australia, until recently, was government owned and operated on a state and interstate level. Improvements in performance of remaining government owned rail systems have taken place following several Productivity Commission reviews and public inquiries since 1989 37. Reforms have been initiated and implemented over this time period culminating with a progress report, “Progress in Rail Reform”, published in 2000 (PC, 13 April 2000).

The progress report states that State railways “still lack a full commercial focus and suffer from inadequate investment”, and suggest that “alternatives to government provision need to

37 See IAC, 1989; EPAC, 1989; IC, 1991 and others (www.pc.gov.au ).

112 be considered” to ensure the future development of the system (PC, 2000, p 125). The AWB Ltd, a participant in the 2000 PC inquiry, supports this statement (PC, 2000, p 126).

The consequence of the lack of specific rail reforms, such as mentioned in the above paragraph, suggests that the quality of rail infrastructure is declining and costs of transportation are increasing. Other forces that have been influential include:

Ø Continued and increasing competition from road transport; Ø Pressure on state government budgets in provision of services to the community (passenger rail concerns deviating funds from freight services); Ø Pressure on freight rates from downstream competition (related more specifically to the minerals and mining sectors); Ø Implementation of the NCP.

(Source: PC, 2001, p 39).

Transportation of grain via road is increasing as “deregulation and microeconomic reform is providing greater opportunity for participation and road transport technology continues to develop” (AWB Ltd, 2001a). South and Western Australia have traditionally been more dependent on road for transport of bulk commodities due to the relative size of the states and the location of receival ports. In SA, specifically, receival ports are much closer to areas of production, which means direct road transport from farm to export terminal is cheaper and more efficient (AWB Ltd, 2001a).

Government funding is a key component to the distribution of freight haulage modes. Most road systems are either state or federally funded, the 1997/98 Commonwealth road expenditure figure was $7 billion for all Federal, State, and Territory roads (DoTRS, ‘Task Outlook’ 2000, p 28). All Australian railtrack is owned and operated by the government via the Australian Railtrack Corporation (ARC) and smaller state based companies, which

113 operates and manages the entire system. Funding provided by the government for rail capital in 1997 was to the order of $250 million over four years (DoTRS, 2000, p 30). A large proportion of Australian railways have been privatised over the last decade and there has been an increasing proportion of private investment into the system (DOTRS, 2000, p 31). It appears however that the funding available for the road system greatly exceeds funding available for rail, which suggests a reason for the greater proportion of freight haulage to utilise road transportation.

BULK HANDLING CORPORATIONS The AWB Ltd has worked closely with Bulk Handling Corporations (BHCs) who co- ordinate grain transport logistics on behalf of marketing companies. BHCs are responsible for grain at storage facilities and during transportation to domestic customers or ports, as well as providing storage prior to shipment. These integrated state based corporations work in close operation with AWB Ltd to ensure provision of “timely and reliable movement of grain from receival points to export terminals and domestic customers, ensuring commitments to customers are met” (AWB Ltd, 2001a).

BHCs facilitate the storage and handling of Australia’s major agricultural bulk commodities, such as wheat. They were previously state based monopolies whose anti-competitive practices have been examined over the last decade 38. Legislation ensured that the BHCs were able to provide average costs across producers and “bundle” all components of the services regardless of whether each service is used (NCP, 2001, Ch 13, p 15). The 2001 NCP Assessment (2001), states that “this monopoly was generally justified by the need to provide growers with equitable access to infrastructure and to avoid duplication” (Ch 13, p 15). Inquiries have focused on the prevention of the misuse of market power due to control of principal grain storage and handling facilities (NCP, 2001, Ch 13, p 15). Reforms have allowed for increased competition by removal of barriers to entry such as the establishment

38 BHCs in South and Western Australia have undergone NCP reviews.

114 of third party rights and price caps on the use and provision of such services (NCP, 2001, Ch 13, p 16). However, the original corporations, due to their vertically integrated structure, have maintained significant control in bulk handling operations.

6.3.2 EXPORT COSTS Exported bulk commodities, requiring shipment by sea, are particularly dependent on a low cost structure to ensure a competitive advantage. The majority of Australian wheat is sold ‘free on board’ (fob) with an increasingly large proportion (30%), of wheat sold as ‘costs, insurance and freight’ (cif), (PC, 1998, pp 152). Cif requires the AWB Ltd to charter a ship to pick up and deliver wheat for export to a specific buyer. The AWB Ltd is responsible for all port authority charges including government levies, stevedoring charges, wharfage, tonnage, navigation charges, berthage and all other loading and delivery costs (PC, 1998, pp 36).

Prior to the privatisation of the AWB Ltd, nearly all wheat was sold fob, with the importer accountable for the product after release from Bulk Handling Corporations (BHCs) and prior to loading39. As a result of fob sales and operating as a statutory marketing authority, the Australian Wheat Board had no contractual relationship with port authorities. There was little or no incentive for shippers to rally for an increase in efficiency as costs were sustained by the buyer (IC, 1993, pp154). Potentially, post deregulation, the AWB Ltd stood to benefit from waterfront reform by decreasing the cost margin included in the comparative price of Australian wheat. The increase in cif sales also provided an incentive for AWB Ltd to demand highly efficient and low cost services.

The Australian maritime industry has undergone significant developments since 1984 with specific reference to shipping, waterfront and port authority reform40. The majority of

39 This process is often referred to as ‘ex spout’ (PC, 1993, p 153, footnote 4). 40 See IC Inquiry Reports, 1993 and 1998 and the Inter-State Commission’s Reports 1984 and 1989.

115

Australia’s port authorities are public agencies (statutory bodies), under State or Territory legislation. Many regional ports deal specifically with Australian bulk exports. Wheat is classified as a bulk export and requires specific bulk handling and loading equipment not available at all ports, hence set up costs for bulk handling ports are often high. As a result many companies, including the AWB Ltd, have specific arrangements with several ports, unlike most other bulk exports (coal, aluminium etc) wheat is handled at over 17 separate ports (IC, 1993, pp8).

The Australian waterfront has generally been seen as a high cost and highly regulated industry in comparison with the industry in other countries (IC, 1993, p 37), (see figure 6.2). In 1989 a Waterfront Industry Reform Authority (WIRA), was established following the Inter-State Commission (ISC), report initiated by the Federal Government. In 1992, reforms were reported to have been “very successful” where “Substantial improvements have been achieved in cost and performance and the industry is now far more competitive” (WIRA, 1992). Specifically, the 1992 review noted that there had been a 20-25% decline in stevedoring charges over the three year period (IC, 1993, pp12). This emphasis on decreasing costs to the shipper has been a priority of waterfront reform over the 1990s.

116

Figure 6.2 International port authority charges for bulk wheat exports

(1992), (A$/t)

Teesport Esperance Brisbane Geraldton Port Lincoln Albany Hamburg Gijon Rotterdam Dunkirk Antwerp Livorno

$0.00 $0.50 $1.00 $1.50 $2.00 $2.50 $3.00

Source: IC, 1993, p 35

The Australian waterfront continues to remain in the reform spotlight. In 1993 the then Industry Commission (now the Productivity Commission), held an initial inquiry into port authority services (IC, 1993). In 1998 the PC released a report investigating international benchmarking standards of the Australian waterfront (PC, 1998) and there is currently a PC inquiry on the harbour towage industries underway (PC, 2002a). The data used in this paper for examining the port costs to the AWB Ltd has been taken from these reports and their various public submissions.

117

6.3.3 DATA – DOMESTIC COSTS Data for the changes in domestic transport costs over the period 1988 till 2000 was calculated solely on rail freight price trends for wheat per tonne 1995-96 to 2000-01. Data for road and BHCs costs was unattainable. Rail data was compiled from a Productivity Commission (2002) report on “Trends in Australian Infrastructure Prices, 1990-91 to 2000- 01”. The data represents the average cost of transporting wheat from storage to ports in each state (PC, 2002a, p 225).

Table 6.3 Real rail freight price trends

Wheat (per tonne) (Index 1996-97 = 100)

NSW VIC QLD SA WA National 1996-97 100 100 100 100 100 100 1997-98 101.3 100.1 100.1 100 98.9 100 1998-99 92.6 99.2 97.0 103.2 96.4 96.2 1999-00 91.4 91.3 93.5 98.4 90.3 92.1 2000-01 78.1 80.1 82.8 93.7 91.3 84.7

(Source: PC estimates based on Australian Bureau of Statistics (ABS, Consumer Price Index, Australia, Cat no. 6401.0); AWB Ltd, Melbourne, personal communication, 8 April 2002 (PC, 2002a, p 225)) 41

41 Note from PC (2002a): The real price index for each State reflects the average cost of transporting wheat from silos to the port. The average is equal to the cost of transporting the grain from each silo, weighted by the tonnage of Export Pool grain moved from that site as a proportion of aggregate State tonnage of AWB Pool Grain to the port for export.

118

The data in Table 6.3, shows a 15.3% decline in national average transport costs of wheat per tonne between the years 1996-97 and 2000-01. If this fall in costs was constant and consistent from 1988-89 one could assume, ceritus paribus, that costs have fallen by 30.6% over the last decade. This figure is in line with an Australasian Railway Association’s Fact Sheet which states:

Efficiency improvements in Australia’s railways have lowered the cost of grain transport (in general) by over 25% since 1990.

(Source: ARA, 2002).

All elasticities and levels of uncertainty are held per the original base case.

DATA – EXPORT COSTS As a result of the material available it has been possible to estimate the port and related government charges associated with the export of bulk wheat, cif, out of various Australian ports. Table 6.4 shows the changes in average per tonne costs in Australian dollars and indicates that during the period 1992 to 2002 there has been a decline in costs of around 9.75%. Costs are in A$ per tonne for 1992 and 2002 for a ship with a gross registered tonnage of 30000 tonnes.

119

Table 6.4 Port authority costs

(A$/t)

1992a 2002b Change $/tn % Change A$/tn A$/tn 1992-2002 1992-2002 PORT Brisbane $2.53 $1.57 $0.96 3.44% Adelaide $2.26 $1.79 $0.47 20.08% Port Lincoln $2.28 $1.71 $0.57 25.00% Esperance $2.48 $2.63 -$0.15 -6.05% Albany $2.23 $2.06 $0.17 7.62% Kwinana $1.06 $1.80 -$0.74 -69.81% Geraldton $2.44 $2.23 $0.21 8.61% Average $2.11 $1.97 $0.21 9.75%

aSource: IC, 1993, Table B8, p 216 bSource: Shipping Australia Ltd, 2002, Attachment C, p 7

Reforms are continuously and simultaneously occurring in both the wheat and the waterfront industries. Hence lower port authority costs are due not only to the AWB Ltd’s attitude to costs, but also to increased efficiency on the waterfront. It is difficult to distinguish which component of the 9.75% decrease over the period 1992-2002 time periods can be accounted for by changes in the AWB Ltd’s corporate structure or by an increase in efficiency in waterfront operations. However, the fact that the AWB Ltd, prior to deregulation, traded in fob contracts suggests that the decline in costs must be accounted for as a result of the AWB Ltd’s change in corporate structure.

120

It is important to note that domestic and overseas costs in the model are representative of total marketing costs and hence a proportion of these total selling costs needs to be allocated specifically to transport costs. The AWB Ltd, report that their ‘Site to Sea’ direct costs42 are approximately 14%, and other marketing costs (pool management fees, insurance and demurrage costs), account for 3% of their National Pool for 2000-01 (AWB, 2001b, p10). Following from this, it can be inferred that transport costs, as a proportion of total costs, are to the order of 82% of total selling costs.

Export costs for the AWB Ltd should also include domestic costs – that is, the transfer of wheat from the receival point to the port. Assume that for exported wheat the internal transport represents 67% of the total transport costs (that is, cd=10 and co=15). Given that domestic costs have declined by 30% over the last decade, the figure used to represent the change in export costs for this time period is 23%.

Modifying these statistics to take into account the proportion of costs allocated to transport (82%), domestic transport costs represent a 25% decline in total domestic selling costs and overseas transport costs represent a 19% decline in total export costs. Note that all elasticities and values of uncertainty are held the constant per the base case.

42 AWB Ltd defines these costs as: “direct costs paid from pool proceeds to service providers involved in the supply chain from up- country receivals sites to bulk wheat shipments, free on board” (AWB, 2001, p 10).

121

6.3.4 RESULTS

Table 6.5 Comparing sales maximisation results when costs are decreased

(export costs down by 19%; domestic costs down by 25%)

Before* Before

co=15 co=12.15

cd=10 cd=7.5

po $32.90 $26.84

pd $27.30 $20.15

QT 188.83 268.88

E(P)Total $3298.32 $3619.37

The results in table 6.5 show a decrease in domestic and export prices when costs have been decreased in both markets. Moreover, these decreases are proportionately in line with the decreases in costs.

122

Table 6.6 Comparing profit maximisation results when costs are decreased

After* After

co=15 co=12.15

cd=10 cd=7.5

po $41.03 $34.43

pd $42.49 $32.18

QT 114.48 159.16

E(P)Total $3473.42 $3809.81

In addition, the results in Table 6.6 show a decline in the prices for both export and domestic wheat, and again these decreases are proportionately in line with the decreases in costs.

Similar to the case of deregulation, these results suggest that if the effects of transport cost reductions appeared in advance of the commercialisation of the AWB being implemented then domestic consumers would have seen this in terms of a decrease in domestic prices until the implementation of commercialisation brought about a price increase. Alternatively, if the impact of transport cost reductions appeared in conjunction with the impact of commercialisation, then no such price cut would be observed. Instead, the extent of the increase in the domestic price associated with commercialisation would simply have been reduced43.

43 It should be noted that, if substantiated, the link between the privatisation of the AWB Ltd and declining overseas and domestic marketing costs could be interpreted as a gain from microeconomic reform.

123

6.4 UNCERTAINTY IN THE INTERNATIONAL ARENA The international trading arena has become increasingly unstable in light of fluctuating exchange rates and general economic uncertainty following September 11th events in 2001 as well as financial upheaval throughout 2002. However, throughout the 1990s there were two periods of global macroeconomic instability, firstly in 1991-93 and then 1998-99 with an average growth rate of 3% (down from 3.5% in the 1980s and 4.5% in the 1970s (IMF, 1999, p 3). The major reason for this instability was currency crises in Mexico, Brazil, Russia and Asia. Many economies remained surprisingly stable throughout this period, namely the USA, Australia, China, India, Ireland, the Netherlands, Norway and Taiwan (IMF, 1999, p4). As a result, the International Monetary Fund (IMF) (1999), believes that:

It is unclear whether macroeconomic instability generally has been increasing. However the mere fact that it has been pervasive may be considered surprising given the general improvement in macroeconomic policies in most countries compared with the two preceding decades

(IMF, 1999, p4).

For the purposes of this study a measure of the change in the level of international uncertainty from 1988/89 to 1998/99 is taken by examining gold prices as a proxy, work by Skousen (1997) suggests that gold is indeed a good proxy for measures of economic stability. It should be noted that there are many means by which to examine international uncertainty, including using crude oil prices, often viewed as highly correlated with gold prices, international commodity figures, such as the Dow Jones Commodity Spot Index or more complicated macroeconomic techniques such as conditional variance of real gross domestic product (Baum et al, 2003). Use of these or other proxies may have alternative implications and the use of gold prices here is purely to demonstrate the direction in which the AWB Ltd’s pricing behaviour may be influenced by international uncertainty.

124

Gold is generally seen as a distinct and relatively stable commodity (less fluctuating than paper currencies), with a “universally acceptable storehouse of value” (Amey, 1998, p 50). Gold is both a commodity and a form of legal tender and it is this dichotomy that enables gold to be used as an indicator for world economic stability – “international political and economic events that may influence the market for gold as a commodity may be outweighed by developments perceived to favor (sic) gold as a medium of exchange” (Amey, 1998, p 50).

The gold standard was established in 1934 and the value of gold remained high until the late 1990s. Gold prices peaked at a historic daily high of US$850 per ounce, 21st January, 1980 following general adverse economic conditions and negative political events specifically in Afghanistan and Iran (Amey, 1998, p 51). Volatility increased during the 1980s as the Japanese began investing in the gold market. The US tightened their monetary policy, computer trading came into force, gold production increased and oil prices fell (Lucas, 1983, p 370 and 1984, p 385). Trends throughout the 1980s were driven by a falling US dollar and a rise in the demand for gold, culminating in the 1987 stock market crash. By 1988 gold prices fell further as positive economic conditions, fuelled by a stronger US dollar, coupled with the withdrawal of the Soviet Union from Afghanistan in 1988, gave way to a period of general political and economic stability.

Throughout the 1990s gold management policies of central banks became increasingly aggressive, boosting gold sales. Financial management flanked with the collapse of the Soviet Union44, the 1992 Recession, and the Gulf War ultimately lead to a further decline in gold prices (Lucas, 1991, p 64-65 and Amey, 1998, p 51). The collapse of the Soviet Union eroded investor confidence in the gold market in the early 1990s and it would have been expected that the effect of a multinational event such as the Gulf War would have caused

44 The USSR are reported to have sold large amounts of gold for hard currency (Amey, 1998, p 51).

125 prices to rise as well. However, it appeared that political stability was deemed to be high in the mid 1990s and the price of gold fell in the years following 1990 (Amey, 1998, p 51).

The gold price rose in 1993 resulting in high stocks and the sell off of scrap gold (Roskill Information Services, 1995). The US dollar was weaker in 1994 and hoarding was further reduced as investors exited the market (Roskill Information Services Ltd, 1995). Prices remained static in the years 1994 to 1996 which was followed by a drastic rise in stocks in 1997 as the Dutch government sold one third of their gold reserves (Gold Fields Mineral Services Limited, 1997, p 5). Alarm spread as this suggested a possible glut if other central banks followed suit (CRU International, 1996, p 19). Banks in the European Union began to sell off their gold stocks in 1997 and 1998 as they prepared for the introduction of the Euro in 2001 (Amey, 1998, p52). Global stability was high in both a political and economic sense and in 1998 the price of gold fell to a low on par with 1979 levels (Amey, 1998, p 52), (See Figure 6.3, below).

126

Figure 6.3 Average annual gold prices

450

400

350

300 Price (US$ per troy ounce)

250 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Year

(US$ per troy ounce)

Source: Amey, 1998, p 53.

Figure 6.3 shows the average annual gold price over the period 1988/89 to 1998/99. Figure 6.4, compares the indexed prices for gold and world wheat and suggests that, with the exception of a reverse trend in 1990/91, that wheat prices tend to mimic gold prices.

These figures indicate that gold prices may be a good proxy for international uncertainty. It is interesting to note that wheat prices appear to follow the same trends.

127

Figure 6.4 Price index for world gold and wheat prices

(base year 1988)

140

120

100

40 Gold Wheat 20 Index Price (base year = 100) 0 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Year

Sources: ABARE, 2002 and Amey, 1998

6.4.1 RESULTS Following the above analysis and using the gold price as a proxy for international uncertainty it is assumed that uncertainty in the overseas market has declined over the last decade by 33%. Costs, elasticities, and domestic uncertainty are held the same as in the base case scenario.

128

Table 6.7 Comparing sales maximisation results when uncertainty in the overseas market decreases

(33%)

Before* Before

Var(uo)=500 Var(uo)=335

Var(ud)=100 Var(ud)=100

po $32.90 $32.84

pd $27.30 $27.44

QT 188.83 188.07

E(P)Total $3298.32 $3301.05

Table 6.7 shows a decline in export prices and a rise in domestic price as international uncertainty decreases. These results follow from the revenue maximiser feeling less at risk generally and therefore willing to bear the increased risk associated with lowering prices to increase expected revenue.

129

Table 6.8 Comparing profit maximisation results when uncertainty in the overseas market decrease

(33%)

After* After

Var(uo)=500 Var(uo)=335

Var(ud)=100 Var(ud)=100

po $41.03 $42.05

pd $42.49 $42.50

QT 114.48 113.08

E(P)Total $3473.42 $3475.02

In Table 6.8 the firm adjusts its prices upwards to reflect the decreased demand uncertainty in the overseas market. In particular, given the firm’s risk aversion, the overseas market is perceived as more attractive, and the price set for that market is raised in the pursuit of increased expected profits even though this also increases the variability of profits.

It follows that as the AWB Ltd has shifted from a revenue maximiser to a profit maximiser changes in the level of international uncertainty can be expected to have had the opposite impact on price setting in the overseas market, with the revenue maximiser avoiding risk with price decreases, and the profit maximiser avoiding risk with price increases.

130

6.5 SENSITIVITY ANALYSIS This section contains an evaluation of the robustness of the hypotheses regarding the impact of commercialisation on the AWB’s domestic and export prices to the contemporaneous occurrence of both deregulation and decreased costs of transport. Specifically, although it was found in sections 6.2 and 6.3 that the contemporaneous occurrence of each of these developments is sufficient only to diminish the extent of the positive impact of commercialisation on domestic and export prices, tables 6.5.1 and 6.5.2 together show that these hypotheses are not robust to the contemporaneous occurrence of both developments. In particular, although with both developments occurring at the same time as commercialisation there continues to be a small increase in the export price, the tables show that the domestic price decreases (ie $27.30 to $26.06).

Moreover, table 6.9 evaluates the sensitivity of this finding to the level of the expected profit constraint on the revenue maximiser, which has been set at 95% of maximum expected profits in the previous analysis, but which is weakened to 90% of maximum expected profits in this table. These results show that a weakening of the expected profit constraint on the revenue maximiser restores the positive impact on prices of commercialisation regardless of the contemporaneous occurrence of deregulation and transport cost decreases (e.g. the domestic price increases from $23.61 to $26.06). It follows that the impact of commercialisation on the AWB’s prices may have been positive or negative depending both on the associated developments of deregulation and cost decreases, and on the weakness of the expected profit constraint on the AWB’s pricing policies prior to commercialisation. In particular, the weaker was this constraint, the more likely it was that both export and domestic prices increased with commercialisation, regardless of the associated developments.

131

Table 6.9 Comparing revenue maximisation results when bd is increased and costs have declined in both markets

Before* Before

bo=1.5

bd=1.3

po $32.90 $24.91

pd $27.30 $18.75

QT 188.83 245.57

E(P)Total $3298.32 $2883.88

Table 6.10 Comparing profit maximisation results when bd is increased and costs have declined in both markets

After* After

bo=1.5

bd=1.3

po $41.03 $34.03

pd $42.49 $26.06

QT 114.48 154.54

E(P)Total $3473.42 $3035.11

132

Table 6.11 Comparing revenue maximisation results with the expected profit constraint set at 90% of maximum expected profits

(Other changes are the same as table 6.9)

Before* Before

bo=1.5

bd=1.3

po $28.85 $22.02

pd $23.61 $16.48

QT 228.60 294.66

E(P)Total $3126.41 $2731.22

6.6 CONCLUSION This chapter has investigated the effects of internal deregulation, transport costs and international uncertainty on the AWB Ltd’s pricing behaviour in the context of the commercialisation of the AWB, where this shift is modelled as a change in its objectives from a revenue to a profit maximiser. The results of the above analyses are evaluated in relation to the developed hypotheses and indicate the impact of recent wheat industry developments on hypothetical prices. In particular, it has been shown that the general effect of commercialisation has been an increase in both domestic and overseas prices. However, during the 1990s in association with commercialisation the Australian wheat industry also experienced deregulation of the domestic market, a decline in wheat transport costs and a decrease in world market uncertainty. Based on the simulation results it has been suggested that because both deregulation and lower transport costs have acted to decrease domestic and export prices, their contemporaneous occurrence with commercialisation will have ameliorated to some extent the price increases associated with commercialisation, and may

133 have even dominated this impact depending on the extent to which the AWB Ltd’s profit constraint was binding on its pricing behaviour prior to commercialisation. In addition, it was found that the commercialisation of the AWB Ltd has resulted in a reversal of the impact of changes in world market uncertainty on the overseas price set by the AWB(I).

Chapter 7 follows with an empirical analysis designed to validate the robustness of the model and the hypotheses developed in last two chapters.

134

CHAPTER 7 EMPIRICAL ANALYSIS

7.1 INTRODUCTION Following from the initial numerical analysis and associated developments in chapters 5 and 6, it is important to examine whether the derived hypotheses for the pricing behaviour of the firm as they shift from a revenue maximiser to a profit maximiser hold when examining actual price data.

Firstly, a detailed account of the data available is made in section 7.2, and then a preliminary empirical analysis into the firm’s pricing behaviour is made in section 7.3. Due to the limited data, a further analysis of the model with reference to the world price is developed in section 7.4 in an attempt to investigate the robustness of this model and the hypotheses made. Finally, conclusions are drawn in section 7.5.

Ideally, a regression analysis would be undertaken using pricing data for the domestic and overseas markets over a significant time period to attempt to capture if any change in pricing behaviour can be attributed to either changes in the firm’s objectives or changes that may have resulted due to other influences on the Australian wheat industry. However, as a result of limited data availability this type of analysis is not possible.

7.2 DATA SET In order to examine the issue of the impact of the changing structure of the AWB over time, ideally time series data on the evolution of prices would be used however, due to regulatory impediments the data set used for this empirical study is limited to publicly available data.

Australian wheat data publicly available is released by the Australian Bureau of Agriculture and Resource Economics (ABARE) in their Australian Commodity Statistics published yearly. This source contains general and non-specific data relating to the supply and

135 disposal of wheat, export price quotations, proportion of wheat exported, an average annual AWB Ltd export quote ($ per tonne), quote price for stock feed, and, total volume of Australian exports by destination and total value.

As a result, a data set has been constructed using the publicly available information found in ABARE’s reports. It should be noted that without such information restrictions many important issues relating to the AWB Ltd and the Australian wheat industry could be researched more fully and assist in further developing the industry in a manner that more closely imitates the reality of the situation.

The data is taken over a thirteen year period, 1988/89 to 2000/01, in order to fit in with the timeframe discussed in the preceding associated developments chapter (chapter 6). The export and domestic price data has been collected from ABARE’s Australian Commodity Statistics (2000, 2001), where export price is given by the AWB Ltd export quote price45 and is supplied in Australian and US dollars per tonne. The domestic price is given by the unit value, in Australian dollars per tonne (ABARE, 2001, p 216). The AWB Ltd’s receivals, in kilotonnes, has been sourced from the AWB Ltd’s ‘Historical Grain Statistics’ (www.awb.com.au).

The raw data (seen in table 7.1), allows some insight into the changes in the AWB Ltd’s pricing behaviour following the (hypothesised) change in the firm’s objectives from revenue to profit maximiser. Application of these prices for the domestic and overseas markets over the last decade allows the true effects of market changes to be analysed and provides insight into the model developed in chapter 5.

45 Average daily asking prices for Australian standard white wheat, free on board, eastern states for the relevant financial year (ABARE, 2001, p 220).

136

Table 7.1 Australian wheat data

(1988/89 - 1998/99)

Year Export Price $A/tn Domestic Price A$/tn Total AWB Receivals Ktn 1988/89 212.00 211.60 12954 1989/90 218.00 195.20 13057 1990/91 161.07 132.00 13382 1991/92 208.77 198.60 8075 1992/93 222.97 178.80 13584 1993/94 223.70 168.90 15123 1994/95 236.97 237.40 7008 1995/96 304.56 260.80 15137 1996/97 264.92 205.80 21866 1997/98 246.22 193.30 14387 1998/99 234.55 178 18033 1999/00 220.72 186.70 21603 2000/01 217.53 218.20 17771

(Source: ABARE, 2000 and 2001).

137

7.3 PRICE RELATIONSHIPS Table 7.2, below shows the ratio over time of the average annual Australian export price in US$ per tonne and the average annual world price in US$ per tonne. This ratio is interesting as it has increased over time, which as expected, indicates some possible changes in pricing behaviour which can be attributed to the deregulation of the AWB Ltd.

Table 7.2 Wheat prices

(US$/tn)

Year World Price Australian Ratio Export Price (WorldP/ExpP) 1988/89 165.75 172.00 1.04 1989/90 160.83 168.00 1.04 1990/91 117.94 126.58 1.07 1991/92 150.96 160.33 1.06 1992/93 141.33 155.67 1.10 1993/94 141.52 153.75 1.09 1994/95 156.42 175.08 1.12 1995/96 215.33 230.42 1.07 1996/97 178.50 206.67 1.16 1997/98 141.75 167 1.18 1998/99 119.80 146.91 1.23 1999/00 113.17 147.52 1.30 2000/01 127.00 149.44 1.18 (Source: ABARE, 2000 and 2001).

138

Figure 7.1 below, represents the data in table 7.2 and clearly shows the relationship between the Australian export price and the world wheat price as well as increasing trend of the ratio of these two prices over the last six years (with the exception of 2000/01).

Figure 7.1 World price and Australian export price for wheat

(US$/tn)

250 1.40

1.20 200 1.00

150 0.80

0.60 100

Price (US$/tn) ExpUS$/tn Ratio (ExpP/WP) WorldPUS$/tn 0.40 50 Ratio expP/WP 0.20

0 0.00 1988-89 1989-90 1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 1998-99 1999-00 2000-01

Year

It is also interesting to look at the ratio of the Australian domestic price to the world price. Table 7.3., below, presents this data which is also examined in graphical form in Figure 7.2. The Australian domestic price for wheat has been converted to US$/tn using the same exchange rate as ABARE for conversion of the Australian export price from A$ to US$. The ratio for domestic price to world price appears to be much more variable than for the export price world price ratio. However, there has also been an increase in this ratio over the

139 last five years. This may also be indicative of the AWB Ltd’s pricing behaviour responding to changes in the objectives of the firm.

Table 7.3 Wheat prices and the price ratio

(US$/tn)

Year World Price Australian Ratio Domestic Price (WP/DomP) 1988/89 165.75 171.68 1.04 1989/90 160.83 150.43 0.94 1990/91 117.94 103.73 0.88 1991/92 150.96 152.52 1.01 1992/93 141.33 124.83 0.88 1993/94 141.52 116.09 0.82 1994/95 156.42 175.40 1.12 1995/96 215.33 197.31 0.92 1996/97 178.50 160.55 0.90 1997/98 141.75 131.11 0.92 1998/99 119.80 111.49 0.94 1999/00 113.17 124.78 1.10 2000/01 127.00 149.90 1.18

(Source: ABARE, 2000 and 2001).

140

Figure 7.2 World wheat price and Australian domestic wheat price

(US$/tn)

250 1.40

1.20 200 1.00

150 0.80

0.60 100 Price (US$/tn)

0.40 Ratio (DomP/WP) DomUS$/tn 50 WorldPUS$/tn 0.20 Ratio DomP/WP

0 0.00 1988-89 1989-90 1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 1998-99 1999-00 2000-01

Year

141

7.3.1 APPLICATION TO THE GENERATED HYPOTHESES The hypotheses generated following the initial numerical analysis outlined in chapter 5 are as follows:

H1: po (P) > po (Re v)

H2: pd (P) > pd (Re v)

H3: QT(Õ) < QT(Re v)

> po (Re v)< pd (Re v)

And in addition:

> po (P)< pd (P)

Table 7.4 shows the ‘actual’ overseas and domestic prices and quantities for the ‘before’ and ‘after’ scenarios given the real data set. Data for the ‘before’ scenario is taken from 1988/89 and data for the ‘after’ scenario from the year 2000/01. The choice of these two points corresponds directly to the start of domestic deregulation in July 1989 and to the corporatisation of the Australian Wheat Board to the AWB Ltd in 1999. This assumes that in the years prior to 1988/89 the former AWB was operating as a statutory marketing authority with government loans and underwriting, aiming to maximise revenues (returns to farmers) and had a monopoly over the domestic wheat market and a monopsony over all Australian wheat for export. From July 1989, the AWB began operating as a profit

142 maximising firm and by 1999 had become the AWB Ltd and had listed on the Australian stock exchange 46.

All prices are presented in Australian dollars and total quantity produced is measured in kilotonnes.

Table 7.4 Price, quantity and expected profit, “before” and “after” - values using real data set

Before After

po $212 $217.53

pd $211.60 $218.20

QT 12954 17771

Examining the actual raw data, we find that the first and second hypotheses hold although the findings for the third hypotheses require further investigation47.

It is important to note that with regard to the domestic price, through the firm’s change in objective from revenue maximiser to profit maximiser, that the results in chapter 5 suggest that if the effects of transport cost reductions appeared in advance of the commercialisation of the AWB being implemented then domestic consumers would have seen this in terms of a decrease in domestic prices. The actual data leads to the conclusion that transport cost

46 See chapter 2 for more detailed description of industry changes during the 1990s 47 Obviously there is little true validity, or testing of the robustness of the model, by comparing the actual raw data to the hypotheses generated by hypothetical data, however, due to data restrictions a more detailed (and preferable regression analysis) is not possible.

143 reductions over the last decade (1989-1999), appeared in conjunction with the impact of commercialisation, as a price cut is observed.

The third hypothesis could be validated by assuming that as a result of growth in international demand for wheat48 there has been growth in the AWB Ltd’s sales. This, coupled with single desk regulations, would suggest that as the AWB Ltd has control over the quantity of wheat produced (and purchased) from farmers as a result of the price premiums achieved in various markets, which are, in theory, passed on to farmers, encouraging increased output and more receivals for the AWB Ltd. It should also be noted that the wheat use in the world market has increased over the last two decades (see figure 7.3).

48 International demand for wheat, specifically in Asian and Middle Eastern markets (two key regions for Australian exports) has been increasing as a result of changes in tastes and preferences as well as rising incomes (see Appendix 1). Forecasters predict an increase in grain consumption over the next few years which should offset the continued increases in global production (Turner et al., 2000, p 31).

144

Figure 7.3 World wheat use

(1978/79 - 2000/01), (Mt)

650

600

550

500

World Wheat Use (Mt) 450

400 1978-79 1980-81 1982-83 1984-85 1986-87 1988-89 1990-91 1992-93 1994-95 1996-97 1998-99 2000-01

Year

(Source: ABARE, 2001, p 221).

Regardless of this preliminary analysis, given the data available there is an alternative method of showing the robustness of the model. World wheat price data, corresponding with the 13 data points, is available. By investigating the behaviour of the world price one can draw inferences of the behaviour of the AWB pre and post privatisation.

7.4 EXTENSION TO THE MODEL Ideally, an investigation of whether the change in real export and domestic prices is consistent with a change in the firm’s objectives would be undertaken. However as an alternative hypothesis in order to test the robustness of the model the price ratios for the

145 profit and revenue maximiser are examined developing H4 (from H1 and H2), following from the results in chapters 5 and 6.

H4: po(P) pd (P) < po (Re v) pd (Re v)

Hypothesis 4 is developed from the export and domestic price ratios for the profit maximiser as compared to the price ratios for the revenue maximiser. Returning to chapters 5 and 6, it can be seen that this hypothesis holds for all scenarios.

Figure 7.4 below shows this price ratio in Australian dollars. The variation of prices suggests that there are likely to be non-systematic effects such as, quality within a wheat class or noise, or confounding effects due to changes in production for example, impacting on prices.

146

Figure 7.4 Australian export and domestic wheat prices and the price ratio

(A$)

350 1.4

300 1.2

250 1

200 0.8

150 0.6

100 0.4 Prices in A$/tonne 50 0.2

0 0

1988-891989-901990-911991-921992-931993-941994-951995-961996-971997-981998-991999-002000-01 Export Price Year Domestic Price Price Ratio

World wheat price (pw) is generally expected to have an ability to affect the pricing behaviour of the AWB Ltd, even assuming they are price setters, and it is important to determine whether the AWB prices are independent of the world price. By examining the price ratio it is hoped that the most significant variation is a result of the variability of the world wheat price over the period 1988/89-2000/01.

The effect of the world price needs to be accounted for prior to drawing any possible conclusions as to the impact of the change of objectives on the pricing behaviour of the AWB Ltd. This is analysed by changing the intercept term (a), (equation (5.5)), of both the ‘before’ and ‘after’ models to see how the AWB’s pricing behaviour responds to shifts in the demand curves at the intercept point.

147

The world price becomes an arbitrary component of the intercept term (a), which allows the demand curves to be shifted as the world price changes. The world price is set at a base case level of 1, where ao = ad = 10 000.

bwi ai = ai pw for i = o, d …(7.1)

Where the new intercept terms are defined as a for both the overseas and domestic markets.

For the base case elasticities of demand are set for the overseas market, bwo = 2 and for the domestic market, bwd = 0.0001. The elasticities are chosen arbitrarily however as the world price is expected to have a very small, if any, impact on the domestic demand function

(rather than price,) bwd is set just above zero.

The original ‘change in objectives’ model is re-run with this additional intercept shifter. A sensitivity style analysis on the response of the overseas and domestic pricing for the AWB as it changes its objectives from a revenue to a profit maximiser is then investigated for a change in world price.

7.4.1 RESULTS The results presented attempt to determine whether the variability of the world price is indeed the main driver for any pricing changes made by the AWB Ltd. If this is the case then it can be assumed that a change in objectives, from sales maximiser to profit maximiser, as a result of government pressure, has had little or no impact on the AWB Ltd’s pricing behaviour.

The ‘base case’ scenario, used in chapter 6, is presented in table 7.4, showing the overseas and domestic prices and price ratios (H4), for the ‘before’ and ‘after’ states using the hypothetical data from the devised from the parameter values presented in section 5.5.

148

Table 7.5 Price, quantity and expected profit, “before” and “after” - the base case scenario world price = 1

Before* After*

po $32.90 $41.03

pd $27.30 $42.49

po/pd 1.21 0.97

World price is then decreased to 0.9 from 1, ceteris paribus, and the model is re-simulated (again using the original hypothetical data), giving the results in table 7.6.

Table 7.6 Price, quantity and expected profit, “before” and “after” - a decline in world price for the Australian Wheat Board

Before After

po $30.00 $40.22

pd $28.86 $42.43

po/pd 1.04 0.95

The results presented in table 7.6 suggest that as the world price declined we would see a fall in the export price and a slight rise, or no change, in the domestic price for the revenue maximiser. The profit maximiser would likely see a decline in the export price, with a small decline, or no change, in domestic price. This follows from general intuition of the behaviour of a firm as a revenue or profit maximiser.

Relating these findings to the actual data proves more complex. Firstly, continuing to use the two data points shown in section 7.3, 1988/89 and 2000/01, the ratio of world price to

149 export price has increased (i.e. 1.04 to 1.18) as the firm has shifted from a revenue maximiser to a profit maximiser, which is inconsistent with the results in table 7.6. The ratio of world price to domestic price has also increased by the same amounts. However, if the change in the price ratio is examined in Figure 7.4 above, it can be seen that the ratio has indeed fallen since the AWB Ltd became a private company under Australian Corporations Law in July 1999. This would suggest that as the AWB Ltd formally adopted the profit maximising aim there has been a change in their pricing behaviour consistent with the change simulated in the model presented in chapter 5. Future research would be able to provide a more conclusive outcome.

7.5 CONCLUSION The extent of any detailed empirical investigation of the model has been severely limited by the lack of available data. Making use of the data that is publicly available has meant that any analysis on the robustness of the model developed in chapter 5 and the results presented in chapters 5 and 6 has to be purely qualitative. An application was made by including the world wheat price as a demand shifter into the original model and attempting to explain that if the pricing behaviour of the revenue and profit maximiser is independent of the world price then the changes in the domestic and overseas prices may be attributable to the change in the AWB’s objectives. The results presented in this chapter are inconclusive although they suggest that the world wheat price appears to be the main influence on the overseas and domestic pricing behaviour of the AWB Ltd as the firm has altered its objectives from a revenue maximiser to a profit maximiser. As noted in section 7.4, further research at a later date would provide more conclusive results as to the degree of robustness of the model developed in chapter 5.

150

CHAPTER 8 CONCLUSION

8.1 INTRODUCTION AND SUMMARY The main aim of this thesis has been to examine contemporary economic and political influences on the Australian wheat industry. The process has been two fold. Firstly, the thesis examined the ability of the AWB International Ltd’s single desk to command market power, and secondly, it examined the impact of a change in objectives of the AWB Ltd since semi-privatisation of the Australian Wheat Board. The principle outcome is a contribution to the academic literature on the Australian wheat industry.

Chapters 2 and 4 provided a review of the current literature and issues including the world wheat industry, the market power and state trading enterprise debates, the Australian wheat industry, Australian microeconomic reform and other relevant literature. Chapter 3 developed the market power analysis and used a traditional price discrimination model based on the Carter-Knetter framework to develop and apply empirically using regression analysis to determine the AWB(I) Ltd’s ability to command price premia. Subsequently, a novel approach was taken to investigate the effects of the microeconomic reform on the pricing behaviour of a price discriminating AWB Ltd. Chapter 5 developed a theoretical model adapted from Fraser (1989) which conceptualised the change of the AWB Ltd’s objectives as a shift from revenue maximization to profit maximization. This model was used to examine the impact of such a change on the pricing policies of a multi-market price-setting firm. Two hypothetical objective functions, a risk averse firm’s price-setting behaviour in an “overseas” and a “domestic” market were analysed, given differing costs of supply, uncertain demand functions and differing price elasticities of demand in each market. The aim was to generate empirically testable hypotheses relating to the impact of a change of objectives on pricing behaviour. In chapter 6, application was made to events that have simultaneously impacted the Australian wheat industry during the last decade, such as the deregulation of the domestic market, changes in transport policy and costs, and international uncertainty. Finally, numerical results and extensions were presented in chapter 7.

151

8.2 KEY FINDINGS AND CONTRIBUTIONS The key findings of this thesis, taken from chapters 3, 5, 6 and 7, are summarised below and each finding’s contribution to the literature is noted.

Chapter 3, presented a traditional price discrimination model of the AWB(I) Ltd and investigated two significant issues. Firstly, the sensitivity of the equilibrium simulation model to changes in the assumption of the functional form of the demand curves was examined. Secondly, an estimation of pricing to market model for the different classes of wheat on disaggregated country data, including an identification of any response in pricing to exchange rate variations was undertaken. The equilibrium simulation model was then re- solved using only estimates of price differentials which are statistically significant, and robust.

The results from these analyses suggest that the functional form of the demand curve can have a large impact on the magnitude of the premiums. Using a linear demand curve increases the estimate of the premium generated by the simulation model, and for the data used here, this can be by a factor of 2-3 times. A priori there is no reason to assume that any specific functional form is correct, which raises some questions about the usefulness of the model if precise estimates of the premia are needed to evaluate the impact of the AWB(I). Alternative functional forms could reduce the estimate of the premia. This leads to a major criticism of the simulation model, and consequently biased results, in the report prepared by the ACG (and other work), in relation to the ability of the AWB(I) to price discriminate.

The pricing to market study, based on the Carter – Knetter Model, and regression analysis indicates statistically significant country specific effects for most classes of wheat traded which suggest some ability to price to market. These effects manifested themselves either as country specific shifters, or a significant relationship between the price being charged and the exchange rate of the importing country. Re-solving the simulation model leads to

152 estimates of premia, which range up to US$2.14 per tonne. For the contracts considered, the average value of the premiums is approximately US$10 million per year. The average premium per tonne across all classes and years is US$ 1.02 per tonne. As a consequence, it may be concluded that the AWB(I) has some ability to set prices for certain types of wheat in some overseas markets. This price discrimination model is the most sophisticated and contemporary pricing model available, to date, using a detailed data set from the AWB Ltd.

Chapter 5 presented a novel model developed to investigate the impact on multi-market pricing behaviour as the objective of a firm is shifted from a revenue-orientated public enterprise to a semi-regulated profit maximiser. Empirically testable hypotheses have been developed through the algebraic and numerical analysis of a risk averse firm’s price-setting behaviour for two different objective functions and given differing costs of supply, uncertainty of demand and differing price elasticities of demand for the firm’s markets.

The model is of a size-orientated price-setting firm operating in multiple and segmented markets. These markets are specified to capture the differences between the AWB Ltd’s overseas and domestic markets. The overseas market is characterised as being a higher cost market, with more elastic and more uncertain demand than the domestic market. With this structure, the model incorporates a “before” and “after” commercialisation pair of objectives for the monopolist, where revenue and profits are the two objectives respectively.

Three empirically-testable hypotheses were generated identifying the likely impact on the AWB Ltd’s overseas and domestic pricing behaviour. In particular, it was suggested that the impact of commercialisation would have been to increase prices in both domestic and overseas markets, with an associated decrease in total sales. This section also showed how the change in objective affects optimal prices when the firm’s markets differ in each respect as well as the combined effect of all differences. This and a further sensitivity analysis of the effect of the firm’s level of risk aversion was conducted which confirmed the robustness

153 of the three hypotheses, but also indicated a set of inconclusive results that will require further attention in an empirical context.

The AWB Ltd’s pricing behaviour has not previously been investigated in this manner, nor has there been detailed research into the impacts of the privatization process either on the AWB Ltd, or other (agricultural) state trading enterprises.

Chapter 6 investigated the effects on the AWB Ltd’s pricing behaviour of policy and other changes, likely to have had an impact on the Australian wheat industry and occurring concurrently with the microeconomic reform programme. Such changes include: internal deregulation, and the levels of transport costs and international uncertainty. Again the shift towards commercial practice was modelled as a change in objectives from a revenue to a profit maximiser.

The results were evaluated in relation to hypotheses developed in chapter 5 and indicate the impact of recent wheat industry developments on hypothetical prices. It was shown that the general effect of commercialisation has been an increase in both domestic and overseas prices. In association with commercialisation the Australian wheat industry, during the 1990s, experienced deregulation of the domestic market, a decline in wheat transport costs and a decrease in world market uncertainty. Simulation results suggest that because both deregulation and lower transport costs have acted to decrease domestic and export prices, their contemporaneous occurrence with commercialisation will have ameliorated, to some extent, the price increases associated with commercialisation, and may have even dominated this impact depending on the extent to which the AWB Ltd’s profit constraint was binding on its pricing behaviour prior to commercialisation. In addition, it was found that the commercialisation of the AWB Ltd has resulted in a reversal of the impact of changes in world market uncertainty on the overseas price set by the AWB(I) Ltd.

154

Chapter 7 made use of the data that is publicly available to present some qualitative support for the results presented in chapters 5 and 6. An application was made by including the world wheat price as a demand shifter into the original model and attempting to explain that if the pricing behaviour of the revenue and profit maximiser is independent of the world price then the changes in the domestic and overseas prices may be attributable to the change in the AWB(I)’s objectives. The results presented in this chapter attempted to validate the model as well as possible, although, it is suggested that the world wheat price appears to be the main influence on the overseas and domestic pricing behaviour of the AWB Ltd as the firm has altered its objectives from a revenue maximiser to a profit maximiser.

8.3 LIMITATIONS AND FURTHER RESEARCH The single largest limitation of this thesis is the availability of relevant data, providing much scope for further research. For the traditional price discrimination model, commercially sensitive data was released by the AWB Ltd, although detailed results cannot be fully reported (chapter 3). Permission for use of this same data set was not granted for any other analysis and hence the robustness of the change in objectives analysis on the pricing behaviour of the AWB Ltd cannot be verified. Publicly available data (used in chapter 7) is highly aggregated, and further, due to the level of aggregation and the contemporary nature of the Australian microeconomic reform process, the number of data observations was limited. The use of this available data was able to help validate the theoretical model as well as possible.

155

APPENDIX 1 AUSTRALIA’S KEY IMPORT MARKETS

Australia exports approximately 30% of their total exports to the Asian market (ABARE, 1999). Wheat is a major component of the Asian diet, specifically noodles, and many Asian nations do not produce enough, if any, wheat to fulfil domestic consumption. Asian markets are key to the revenue of Australian wheat farmers. Figure A.1 below, shows the trends in Australian wheat exports to primary Asian markets during the 1990s. Since the mid 1990s wheat exports rose and look to be continuing this trend. The dramatic blow out and tapering off of Australian exports into Indonesia in 1997 follows the rise of the Newly Industrialised (Asian) Countries (NICs), and the 1997 Asian Crisis.

Carter and Lyons (1996) in their paper “The Economics of Single Desk Selling of Western Canadian Grain”, classify the wheat market into two main categories, “a small high quality, high priced market” (Ch 1, p 2), like the UK or Japan and a much larger “lower quality and lower priced market” (Ch 1, p 2) in emerging markets such as China, Iran and Egypt. They also comment that there is growth in the middle ground for medium quality wheats to nations in Asia, such as Indonesia, Malaysia and South Korea as a result of higher incomes and changes in tastes and preferences.

156

Figure Appendix.1 Australian wheat and wheat flour exports (Kt), 1991–1999

Australian Wheat & Flour Exports, 1991 - 1999

3000

2500

2000 Indonesia SthKorea 1500 Malaysia

Volume (kt) 1000 Pakistan

500

0 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 1998-99

Years

(Source: ABARE, 1999)

INDONESIA Indonesia does not produce any wheat and hence requires 100% imports for consumption. Indonesia imported over 3.5 million tonnes in the 1998/99 period. In 1996 Indonesians spent over 23% of their monthly earnings on cereals, more than all other food groups. Generally Indonesian’s demand for wheat should continue to increase in line with populations growth, although it should be noted that overall consumption of food has fallen 6 percentage points between 1987 and 1996 in favour of non-food items. Such activity indicates an increase in per capita disposable income.

157

Indonesia has been an important market for Australia since the 1980s. In 1986, Australia exported just over 500 000 tonnes of wheat (AWB(I), 2000), and this figure more than doubled by 1998/99 to 1.4 million tonnes of wheat and wheat flour (ABARE, 1999).

Economic improvement in Indonesia has meant a resurgence of wheat imports (up 13% in 1999 to 3 million megatonnes). Consumption of wheat based foods, specifically noodles, is growing (Johnson and Niniek, 2000). Australia supplies over 55% of the market (mainly Australian Soft White wheat), with Canada holding a 30% share (Johnson and Niniek, 2000). The US is set to increase their exports to the region following implementation of credit guarantee programmes and other relief programmes, such as US$15 million under PL- 480 food aid programme in 1999 (Johnson and Niniek, 2000). Consequently, the US increased their market share to 9% in 1999, up from 1.5% in the previous year (Johnson and Niniek, 2000). It should also be noted that the US has pitched 50 000 metric tonnes of wheat to be donated to Indonesia by the end of financial year, 2000. This wheat has been donated under Section 416(b). A fifth of this wheat is designated for use in a Global Food for Education scheme and the remaining wheat for use as “low grade noodle wheat” to be sold internally at lower than world prices (FAS PR 0355-00 and PR0361-00, September, 2000).

Prior to 1998 the Indonesian government acted as a single desk purchaser of wheat imports, through the procurement agency “BULOG”. During 1998 liberalisation occurred as a result of internal and external pressures. Australia still deals with a government owned flour mill, Bogasari, which previously held about an 80% share of the Indonesian flour market (JISG, 2000, p 89). This huge share of the internal market suggests that Bogasari is still likely to be in a strong position for procuring imports and AWB(I)’s involvement with the mill is likely to be a reason for Australia’s strong market position (58% market share, 1998/99). Another reason for strong Australian market share freight adavantage, although a possible government to government link (Bogasari and AWB(I)), could not be entirely ruled out.

158

REPUBLIC OF SOUTH KOREA Korea has imported an average of 4 million tonnes of wheat per year over the last 10 years, of which approximately 1 million tonnes is from Australia. South Korea has a small domestic production of wheat, on average 2 million tonnes per year, but this figure has been declining. “Since the liberalization of the wheat market in 1984, Korea has been almost totally dependent on imports” (South Korean Ministry of Agriculture and Fisheries, 2000).

Prior to 1984, Korea had high tariff rates, quotas and non-tariff barriers on agricultural goods. They faced a considerable amount of external pressures as a member of the GATT, especially from Canada, the US and Australia. In 1984 Korea began decreasing barriers to trade and since the creation of the WTO in 1995 further liberalization has occurred.

MALAYSIA Malaysia does not produce wheat and hence relies solely on imports to fulfil its demand. The 1997 Asian Currency Crisis caused some concern for importers to Malaysia, although generally Australia faired well at the expense of nations such as the US, due to a smaller exchange rate effect, “giving them an additional margin to undercut US products” (FAS, 1998a).

The immediate effects of the Asian Crisis were met with a slackening in demand for wheat because of Malaysia’s reliance on imports (FAS, 1998a). However, increased economic growth since the crisis is again fuelling changing consumption patterns as Malaysians shift from rice to wheat based products such as noodles and baked goods. Long term conditions for wheat importers seem positive (FAS, 1998a).

Australia has held up to a 70% market share for wheat imports in Malaysia since the 1995/96 season (FAS, 1998a). The US blame their lagging market share on the 1997 currency crisis and Australia’s export monopoly claiming that Australia “can target specific markets and undercut prices of non-monopoly markets” (FAS, 1998a). The US were planning to

159 counteract Australia’s share through increased development of wheat classes centred specifically on wheat for use in noodles (FAS, 1998a).

INDIA AND PAKISTAN The Indian wheat market is highly subsidised to encourage local producers and is set at about US$134 per tonne with above average wheat stocks of approximately 16 million tonnes in 2000 (Govindan, 2000). These factors suggest that India’s wheat imports will be low in the short run as they will concentrate on using locally produced wheat and depleting stocks. However, with an ever increasing population and sporadic demand there is potential for exporters to gain a significant share in the Indian market.

Pakistan produces approximately 15 million tonnes per year of wheat (ABARE, 1999), but nonetheless imports another 4 million tonnes every year to fulfill their demand for wheat and wheat flour. Australia began exporting to Pakistan in the early 1980s. During this time trade was volatile and it was not until 1996/7 that Australia began exporting large volumes of wheat. Since 1996 Australia has exported about 1 million tonnes per year (AWB(I), 2000). The Pakistani government offers assistance to their wheat farmers in order to encourage domestic production. Farmers currently receive about US$4.80 subsidy per 40kg of wheat produced (1999/2000). The Pakistani government sees this as a saving of up to US$250 million that would otherwise have been spent on wheat imports.

CHINA China has potential to be a growing Asian market for wheat, however, their push to raise domestic production in line with a self sufficiency and food security argument (FAS, 1998b), exhibits an unwillingness to import. The Chinese refuse to be dependent on any one supplier for strategic reasons and hence competition is fierce (FAS, 1998b). China is the world’s largest producer of wheat and consumption is high in line with increases in economic growth, forecasted at 8.8% for 2000 (FAS, 1998b). Economic growth generally

160 encourages a shift in dietary patterns and boosts demand for higher quality wheats for pastry manufacturing as well as for feed for livestock (FAS, 1998b).

Regardless of China’s aims to become self reliant, it is likely they will face production shortfalls, as a result of competition for land use, and will be forced to import up to 5% of their total demand for wheat (FAS, 1998b). As a result there are possibilities for nations such as Australia to continue to export to China. The developments of 1999/2000 in relation to China’s entry into the WTO has sparked uncertainty in the expected import quantities and origin, though imports are set to rise with a forecasted shortage of cropping area and stocks in China by the 2001 (Wade and Zhang, 2000, p 1).

China has three agricultural support aims which are met through price controls, and market manipulation (FAS, 1998b).

Provide stable farm income; Increase production, and Maintain low grain prices in urban areas via adequate supply.

(FAS, 1998b).

Many of the Chinese agricultural policies are believed to inhibit free trade including trade prohibitive quotas and strict import license arrangements as well as dubious quality controls:

“China has implemented sanitary and phytosanitary barriers that the US and other countries believe are not scientifically sound in an apparent attempt to exclude imports of certain commodities”

(FAS, 1998b).

161

With China’s ascension to the WTO in 2002 it is hoped that these protectionist policies will be relaxed.

JAPAN Japan is generally regarded as a quality conscious and price insensitive market and is also seen as the most competitive market for wheat exports. Australia competes intensively with the US and Canada in the Japanese market. As a result many previous studies have focused on the Japanese import market with ambiguous conclusions. Some economists believe that the Japanese control the flow of wheat into their country and it is the Japanese Food Agency (JFA) which holds market power (Stiegert and Blanc, 1997 and Love and Murniningtyas, 1992).

CONCLUSION Asia is the key importer of Australian wheat will remain an important market for Australian wheat in the future. These nations have also begun to enjoy higher national income levels and are beginning to consume more western style foods, bolstering their demand for wheat and wheat products. With this more Westernized diet a higher proportion of meat is also being consumed, and there is much potential for feed wheats to be exported to the region. The AWB(I) needs to continue to focus on end use requirements and other market possibilities in the region.

162

APPENDIX 2 REGRESSION ANALYSIS RESULTS

Estimation results for the pricing to market study (Chapter 3), by classes of wheat (Class I, Class II etc.), are presented below. mon_1, mon_2 etc. are monthly dummy variables, where mon_1 equals 1 in January year 1, and zero otherwise: mon_2 =1 in February year 1, and so on. The country names have been replaced by (random) codes to maintain confidentiality.

ERi denotes the exchange rate variable for country i. Country labels are consistent within equations (i.e. if a country specific dummy and exchange rate are both included, they can be identified as such). Only those that were significant at the 10% level were retained in the equation. Igrade_n are grade dummies.

163

Class I

Source | SS df MS Number of obs = 720 ------+------F( 55, 664) = 122.45 Model | 19.7788474 55 .359615408 Prob > F = 0.0000 Residual | 1.95012576 664 .002936936 R-squared = 0.9103 ------+------Adj R-squared = 0.9028 Total | 21.7289732 719 .030221103 Root MSE = .05419

------| Coef. Std. Err. t P>|t| [95% Conf. Interval] ------+------mon_8 | .406418 .0555686 7.314 0.000 .2973066 .5155294 mon_9 | .3905952 .0206329 18.931 0.000 .3500816 .4311088 mon_10 | .3732462 .0177262 21.056 0.000 .3384401 .4080524 mon_11 | .3375369 .0155175 21.752 0.000 .3070676 .3680063 mon_12 | .3119969 .0166241 18.768 0.000 .2793547 .3446391 mon_13 | .313542 .0150685 20.808 0.000 .2839543 .3431297 mon_14 | .2981023 .0150156 19.853 0.000 .2686186 .327586 mon_15 | .3112515 .0162245 19.184 0.000 .2793939 .3431091 mon_16 | .3270485 .0172466 18.963 0.000 .2931841 .360913 mon_17 | .2965181 .0176786 16.773 0.000 .2618054 .3312307 mon_18 | .2268798 .0179925 12.610 0.000 .1915507 .2622089 mon_19 | .1279926 .0189919 6.739 0.000 .0907011 .1652841 mon_20 | .2460553 .0160834 15.299 0.000 .2144748 .2776358 mon_21 | .2252457 .0145546 15.476 0.000 .1966671 .2538 243 mon_22 | .2240219 .014738 15.200 0.000 .1950832 .2529606 mon_23 | .2005618 .0163809 12.244 0.000 .1683972 .2327264 mon_24 | .203911 .0200001 10.195 0.000 .1646399 .243182 mon_25 | .147765 .0158391 9.329 0.000 .1166642 .1788658 mon_26 | .169907 .0142846 11.894 0.000 .1418586 .1979554 mon_27 | .1606174 .0161159 9.966 0.000 .1289732 .1922616 mon_28 | .1002941 .0163189 6.146 0.000 .0682513 .1323369 mon_29 | .0851831 .020189 4.219 0.000 .0455411 .1248251 mon_30 | .0278557 .0140208 1.987 0.047 .0003253 .0553861 mon_31 | -.0586625 .014849 -3.951 0.000 -.0878191 -.0295059 mon_32 | -.0614806 .0160218 -3.837 0.000 -.0929401 -.0300212 mon_33 | -.0233326 .0163327 -1.429 0.154 -.0554024 .0087373 mon_34 | .1196887 .0159856 7.487 0.000 .0883003 .1510772 mon_35 | .0720553 .0139812 5.154 0.000 .0446027 .099508 mon_36 | .0574368 .0149734 3.836 0.000 .0280359 .0868377 mon_37 | .0749884 .0149337 5.021 0.000 .0456654 .1043114 mon_38 | .0295841 .0175475 1.686 0.092 -.0048712 .0640395 mon_39 | .0305058 .0196496 1.552 0.121 -.0080771 .0690886

164

mon_41 | -.0118855 .0141238 -0.842 0.400 -.0396183 .0158472 mon_42 | -.0301162 .0162269 -1.856 0.064 -.0619784 .001746 mon_43 | -.0501712 .0153221 -3.274 0.001 -.0802567 -.0200857 mon_44 | .0002714 .0158 0.017 0.986 -.0307525 .0312954 mon_45 | -.0064432 .0196886 -0.327 0.74 4 -.0451027 .0322162 mon_46 | .0190246 .0207572 0.917 0.360 -.021733 .0597822 mon_47 | -.0235478 .0406213 -0.580 0.562 -.1033095 .0562139 Igrad_5 | -.1051664 .0173396 -6.065 0.000 -.1392135 -.0711193 Igrad_27| .0985968 .0098541 10.006 0.000 .0792478 .1179457 Igrad_28 | -.027752 .0115891 -2.395 0.017 -.0505077 -.0049962 Igrad_33 | .091413 .0576433 1.586 0.113 -.0217722 .2045982 Igrad_49 | .0721762 .0131945 5.470 0.000 .0462681 .0980842 a | -.1620551 .0304469 -5.323 0.000 -.2218389 -.1022713 b | -.0666687 .0264711 -2.519 0.012 -.1186459 -.0146915 c | -.1499754 .0307483 -4.878 0.000 -.2103511 -.0895998 d | .1162656 .0260364 4.466 0.000 .065142 .1673892 e | -.002516 .026277 -0.096 0.924 -.054112 .0490799 f | .0629846 .0304788 2.067 0.039 .0031383 .122831 g | .0678996 .038071 1.784 0.075 -.0068544 .1426536 ER(c) | -.1988363 .0943494 -2.107 0.035 -.3840955 -.0135772 ER(d) | .0639175 .0207601 3.079 0.002 .0231543 .1046808 ER(g) | .1517205 .0800433 1.895 0.058 -.005448 .3088889 ER(h) | .137797 .0697258 1.976 0.049 .0008873 .2747067 _cons | 4.864474 .0275744 176.413 0.000 4.810331 4.918618 ------

165

Class II

Source | SS df MS Number of obs = 317 ------+------F( 52, 264) = 30.17 Model | 4.78725185 52 .092062536 Prob > F = 0.0000 Residual | .805653343 264 .003051717 R-squared = 0.8560 ------+------Adj R-squared = 0.8276 Total | 5.59290519 316 .017699067 Root MSE = .05524

------lnfob2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ------+------mon_6 | .5756256 .0594234 9.687 0.000 .4586215 .6926297 mon_8 | .4525873 .0594234 7.616 0.000 .3355832 .5695914 mon_9 | .426065 .0599425 7.108 0.000 .3080388 .5440912 mon_11 | .2808813 .0253021 11.101 0.000 .2310616 .3307009 mon_12 | .2650839 .028124 9.426 0.000 .2097079 .3204598 mon_13 | .2832066 .0269385 10.513 0.000 .230165 .3362482 mon_14 | .2940959 .0314041 9.365 0.000 .2322616 .3559303 mon_15 | .3020269 .0308743 9.782 0.000 .2412357 .362818 mon_16 | .2750887 .0318732 8.631 0.000 .2123306 .3378468 mon_17 | .287677 .0286217 10.051 0.000 .231321 .3440329 mon_18 | .2444156 .0290878 8.403 0.000 .1871421 .3016892 mon_19 | .2114895 .0335963 6.295 0.000 .1453386 .2776404 mon_20 | .2617054 .0260423 10.049 0.000 .2104284 .3129825 mon_21 | .2402768 .0280318 8.572 0.000 .1850825 .2954711 mon_22 | .1754197 .0303805 5.774 0.000 .1156007 .2352387 mon_23 | .2455398 .0329107 7.461 0.000 .180739 .3103406 mon_24 | .1605441 .0241991 6.634 0.000 .1128963 .2081919 mon_25 | .1506117 .0254434 5.919 0.000 .100514 .2007094 mon_26 | .1557679 .0287861 5.411 0.000 .0990885 .2124474 mon_27 | .1901525 .0273456 6.954 0.000 .1363094 .2439957 mon_28 | .0836527 .0369742 2.262 0.024 .0108507 .1564546 mon_29 | .1068299 .0225772 4.732 0.000 .0623756 .1512843 mon_30 | -.0078625 .0271546 -0.290 0.772 -.0613296 .0456045 mon_31 | -.0261385 .0276541 -0.945 0.345 -.0805891 .0283121 mon_32 | -.0508359 .0248973 -2.042 0.042 -.0998586 -.0018133 mon_33 | -.0455392 .0305925 -1.489 0.138 -.1057757 .0146972 mon_34 | .0916623 .024701 3.711 0.000 .0430263 .1402983 mon_35 | .0511597 .0368411 1.389 0.166 -.0213802 .1236995 mon_36 | .020479 .0236879 0.865 0.388 -.0261622 .0671203 mon_37 | .0367929 .0240942 1.527 0.128 -.0106483 .0842342

166

mon_38 | -.0003753 .0243917 -0.015 0.988 -.0484023 .0476516 mon_39 | .0181369 .0282431 0.642 0.521 -.0374735 .0737472 mon_41 | -.0119588 .0247717 -0.483 0.630 -.060734 .0368164 mon_42 | -.0277251 .0232625 -1.192 0.234 -.0735288 .0180786 mon_43 | -.035115 .021057 -1.668 0.097 -.0765761 .0063461 mon_44 | -.0254638 .0248357 -1.025 0.306 -.074365 .0234375 mon_45 | -.0065984 .0228381 -0.289 0.773 -.0515665 .0383696 mon_46 | -.0214374 .0300911 -0.712 0.477 -.0806864 .0378116 mon_47 | -.0870127 .058051 -1.499 0.135 -.2013145 .0272892 Igrad_17 | .0330407 .0592749 0.557 0.578 -.0836711 .1497524 Igrad_18 | .0544625 .0638456 0.853 0.394 -.0712488 .1801739 Igrad_19 | .0161575 .0666927 0.242 0.809 -.1151597 .1474748 Igrad_36 | .1050691 .0612103 1.717 0.087 -.0154533 .2255915 a | .0055241 .0185415 0.298 0.766 -.0309839 .042032 b | -.0329608 .0193263 -1.705 0.089 -.0710142 .0050926 c | -.018039 .0186054 -0.970 0.333 -.0546729 .0185949 d | .0104088 .0229405 0.454 0.650 -.0347609 .0555785 e | .0192677 .0195479 0.986 0.325 -.019222 .0577574 f | -.0380485 .0162642 -2.339 0.020 -.0700725 -.0060245 g | .0726464 .0236823 3.068 0.002 .0260163 .1192766 ER(h) | .1403886 .0561554 2.500 0.013 .029819 .2509581 ER(i) | -.1432606 .0404815 -3.539 0.000 -.2229682 -.063553 _cons | 4.800269 .0644927 74.431 0.000 4.673283 4.927255 ------

167

Class III

Source | SS df MS Number of obs = 254 ------+------F( 48, 205) = 26.77 Model | 2.32178824 48 .048370588 Prob > F = 0.0000 Residual | .370369666 205 .001806681 R-squared = 0.8624 ------+------Adj R-squared = 0.8302 Total | 2.69215791 253 .01064094 Root MSE = .04251 ------| Coef. Std. Err. t P>|t| [95% Conf. Interval] ------+------mon_1 | .1710664 .0471508 3.628 0.000 .0781037 .2640291 mon_10 | .2807925 .0236556 11.870 0.000 .2341529 .327432 mon_11 | .1842221 .025509 7.222 0.000 .1339285 .2345157 mon_12 | .1890767 .0244798 7.724 0.000 .1408123 .2373412 mon_13 | .0953195 .031794 2.998 0.003 .0326344 .1580045 mon_14 | .1744453 .0226949 7.687 0.000 .1296999 .2191906 mon_15 | .1603395 .0316517 5.066 0.000 .0979348 .2227442 mon_16 | .193765 .0289168 6.701 0.000 .1367525 .2507776 mon_17 | .1647387 .0235605 6.992 0.000 .1182869 .2111906 mon_18 | .1618365 .0312762 5.174 0.000 .1001723 .2235007 mon_19 | .1349798 .0240858 5.604 0.000 .0874922 .1824675 mon_20 | .1783778 .0238868 7.468 0.000 .1312824 .2254731 mon_21 | .1686935 .0221986 7.599 0.000 .1249266 .2124603 mon_22 | .1530538 .0238806 6.409 0.000 .1059708 .2001368 mon_23 | .1455071 .0263438 5.523 0.000 .0935676 .1974467 mon_24 | .1377623 .0234441 5.876 0.000 .0915399 .1839847 mon_25 | .1439165 .0292597 4.919 0.000 .086228 .201605 mon_26 | .0975545 .0230993 4.223 0.000 .0520119 .1430972 mon_27 | .1142758 .0259086 4.411 0.000 .0631944 .1653573 mon_28 | .0598607 .0268826 2.227 0.027 .0068589 .1128625 mon_29 | .0610069 .0229814 2.655 0.009 .0156967 .1063171 mon_30 | .0386755 .0259553 1.490 0.138 -.012498 .089849 mon_31 | .0557764 .0246425 2.263 0.025 .0071912 .1043617 mon_32 | -.028147 .0285225 -0.987 0.325 -.084382 .028088 mon_33 | -.0490375 .027413 -1.789 0.075 -.1030851 .0050101 mon_34 | .01852 .0364979 0.507 0.612 -.0534394 .0904794 mon_35 | .0259532 .0218853 1.186 0.237 -.017196 .0691024 mon_36 | .0604134 .0274262 2.203 0.029 .0063398 .114487 mon_37 | .0195908 .0245479 0.798 0.426 -.0288079 .0679894 mon_38 | .0002858 .0243027 0.012 0.991 -.0476294 .0482011 mon_39 | .0208189 .0285651 0.729 0.467 -.0355001 .0771379 mon_41 | .0150142 .0466519 0.322 0.748 -.0769648 .1069933 mon_42 | .0061899 .0249663 0.248 0.804 -.0430338 .0554136

168

mon_43 | .0004154 .0466519 0.009 0.993 -.0915637 .0923944 mon_44 | .0631168 .024404 2.586 0.010 .0150019 .1112318 mon_45 | .0464006 .0466519 0.995 0.321 -.0455785 .1383796 mon_46 | -.020718 .0471508 -0.439 0.661 -.1136806 .0722447 Igrad_39 | -.3321966 .0758578 -4.379 0.000 -.4817581 -.182635 Igrad_41 | -.1898071 .0445305 -4.262 0.000 -.2776035 -.1020107 Igrad_42 | -.1584575 .0460903 -3.438 0.001 -.2493293 -.0675857 Igrad_43 | -.1711864 .0459977 -3.722 0.000 -.2618756 -.0804971 Igrad_44 | -.1817027 .0562708 -3.229 0.001 -.2926464 -.070759 Igrad_45 | -.1199011 .0641422 -1.869 0.063 -.246364 .0065618 a | -.0201174 .0113799 -1.768 0.079 -.042554 .0023193 b | -.0236958 .0186579 -1.270 0.206 -.0604817 .0130902 c | .1107188 .0095572 11.585 0.000 .0918758 .1295619 d | .0481487 .0164575 2.926 0.004 .0157011 .0805964 ER(d) | .1435467 .0397056 3.615 0.000 .0652629 .2218305 _cons | 5.214471 .0490942 106.213 0.000 5.117677 5.311266 ------

169

Class IV

Source | SS df MS Number of obs = 382 ------+------F( 56, 325) = 43.63 Model | 4.78369851 56 .085423188 Prob > F = 0.0000 Residual | .636313986 325 .001957889 R-squared = 0.8826 ------+------Adj R-squared = 0.8624 Total | 5.42001249 381 .014225755 Root MSE = .04425

------lnfob2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ------+------mon_8 | .4198663 .0471679 8.902 0.000 .3270733 .5126594 mon_9 | .3356778 .0269883 12.438 0.000 .282584 .3887717 mon_11 | .2860678 .0209302 13.668 0.000 .244892 .3272436 mon_12 | .2571547 .0201529 12.760 0.000 .2175081 .2968012 mon_13 | .2697694 .0206476 13.065 0.000 .2291496 .3103892 mon_14 | .2880666 .0197021 14.621 0.000 .2493069 .3268263 mon_15 | .2908099 .0199147 14.603 0.000 .2516319 .3299879 mon_16 | .3178498 .0220196 14.435 0.000 .2745309 .3611687 mon_17 | .2848842 .0261357 10.900 0.000 .2334677 .3363007 mon_18 | .1797509 .0171044 10.509 0.000 .1461015 .2134002 mon_19 | .1486564 .0191224 7.774 0.000 .111037 .1862757 mon_20 | .1917436 .0181434 10.568 0.000 .1560503 .2274369 mon_21 | .1945779 .019836 9.809 0.000 .1555548 .233601 mon_22 | .2135076 .0199404 10.707 0.000 .1742791 .2527362 mon_23 | .2045346 .018427 11.100 0.000 .1682833 .2407859 mon_24 | .1824812 .0234584 7.779 0.000 .1363318 .2286306 mon_25 | .1435416 .0169394 8.474 0.000 .1102168 .1768663 mon_26 | .1565449 .0177495 8.820 0.000 .1216264 .1914634 mon_27 | .1399403 .0201599 6.942 0.000 .1002799 .1796008 mon_28 | .1356756 .020301 6.683 0.000 .0957376 .1756135 mon_29 | .0831562 .0237479 3.502 0.001 .0364373 .1298751 mon_30 | .026001 .0188244 1.381 0.168 -.011032 .0630339 mon_31 | .0279725 .017599 1.589 0.113 -.0066498 .0625948 mon_32 | -.0311616 .0165493 -1.883 0.061 -.0637188 .0013956 mon_33 | -.0102932 .017167 -0.600 0.549 -.0440656 .0234792 mon_34 | .106778 .0164911 6.475 0.000 .0743353 .1392207 mon_35 | .1111878 .0172 99 6.427 0.000 .0771556 .14522 mon_36 | .0779542 .0167994 4.640 0.000 .0449049 .1110036 mon_37 | .0906587 .0162275 5.587 0.000 .0587344 .122583 mon_38 | .0216966 .0156747 1.384 0.167 -.0091402 .0525333 mon_39 | .0358161 .0162665 2.202 0.028 .0038153 .067817 mon_41 | -.0031814 .0153774 -0.207 0.836 -.0334332 .0270704

170

mon_42 | .0256627 .0190895 1.344 0.180 -.0118919 .0632174 mon_43 | -.0452472 .0194232 -2.330 0.020 -.0834584 -.0070361 mon_44 | .0053513 .0332099 0.161 0.872 -.0599823 .0706849 mon_45 | .0193543 .021403 0.904 0.367 -.0227516 .0614602 mon_46 | -.0090982 .02534 -0.359 0.720 -.0589493 .0407529 Igrad_8 | .0095183 .0278789 0.341 0.733 -.0453275 .0643642 Igrad_9 | -.0260319 .0238921 -1.090 0.277 -.0730347 .0209708 Igrad_10 | -.0282055 .0224417 -1.257 0.210 -.0723548 .0159438 Igrad_12 | .0483625 .0243931 1.983 0.048 .0003742 .0963508 Igrad_13 | -.0377186 .0285001 -1.323 0.187 -.0937867 .0183494 a | .0369506 .0226611 1.631 0.104 -.0076302 .0815315 b | .1076891 .0241885 4.452 0.000 .0601034 .1552749 c | .0217067 .0211385 1.027 0.305 -.0198789 .0632923 d | .0630156 .0209733 3.005 0.003 .02175 5 .1042762 e | -.0657857 .0337306 -1.950 0.052 -.1321435 .000572 f | .0746818 .0239322 3.121 0.002 .0276003 .1217633 g | .0803176 .0426591 1.883 0.061 -.0036052 .1642404 h | .0411461 .021664 1.899 0.058 -.0014732 .0837655 Inc2_32 | .0252926 .0248185 1.019 0.309 -.0235326 .0741179 ER(i) | .024798 .0137385 1.805 0.072 -.0022296 .0518257 ER(j) | .0983952 .0281922 3.490 0.001 .042933 .1538573 ER(e) | -3.779899 1.813894 -2.084 0.038 -7.348354 -.2114438 ER(g) | .1434239 .0837823 1.712 0.088 -.0214002 .308248 ER(k) | -.113807 .049054 -2.320 0.021 -.2103105 -.0173035 _cons | 4.873773 .032605 149.479 0.000 4.809629 4.937916 ------

171

Class V

Source | SS df MS Number of obs = 88 ------+------F( 39, 48) = 22.15 Model | 1.94731657 39 .049931194 Prob > F = 0.0000 Residual | .108206591 48 .002254304 R-squared = 0.9474 ------+------Adj R-squared = 0.9046 Total | 2.05552316 87 .023626703 Root MSE = .04748

------lnfob2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ------+------mon_12 | .090145 .0498586 1.808 0.077 -.0101025 .1903925 mon_13 | .0781931 .0516092 1.515 0.136 -.0255742 .1819604 mon_14 | .1442863 .0614841 2.347 0.023 .0206643 .2679083 mon_17 | .1894037 .0476394 3.976 0.000 .0936182 .2851892 mon_18 | .0465241 .0496717 0.937 0.354 -.0533476 .1463957 mon_19 | .0813793 .0505488 1.610 0.114 -.0202557 .1830144 mon_21 | .089212 .0488991 1.824 0.074 -.0091063 .1875303 mon_23 | .081576 .0462967 1.762 0.084 -.011509 7 .1746618 mon_24 | .0889693 .0628305 1.416 0.163 -.0373599 .2152984 mon_25 | -.0206436 .0614841 -0.336 0.739 -.1442656 .1029784 mon_27 | .0199023 .0614841 0.324 0.748 -.1037197 .1435243 mon_28 | -.0616812 .0614841 -1.003 0.321 -.1853032 .0619408 mon_29 | -.0466036 .0441741 -1.055 0.297 -.1354215 .0422143 mon_30 | -.0553306 .0614951 -0.900 0.373 -.1789748 .0683137 mon_31 | -.2332526 .0624888 -3.733 0.001 -.3588947 -.1076104 mon_32 | -.1246241 .0381589 -3.266 0.002 -.2013478 -.0479004 mon_33 | -.1970118 .0428165 -4.601 0.000 -.2831002 -.1109235 mon_35 | -.0415531 .0369538 -1.124 0.266 -.1158536 .0327474 mon_36 | -.0392402 .0520374 -0.754 0.454 -.1438684 .0653881 mon_37 | -.0241993 .0357759 -0.676 0.502 -.0961316 .047733 mon_38 | -.0454516 .0444027 -1.024 0.311 -.1347292 .043826 mon_39 | -.0211349 .0438007 -0.483 0.632 -.1092022 .0669323 mon_41 | .0159318 .0512905 0.311 0.757 -.0871945 .1190582 mon_43 | .0922499 .0671462 1.374 0.176 -.0427565 .2272563 mon_44 | -.0897103 .0409882 -2.189 0.034 -.1721226 -.0072979 mon_45 | -.0472124 .0527279 -0.895 0.375 -.1532289 .0588042 mon_46 | .005294 .0614951 0.086 0.932 -.1183502 .1289383 mon_47 | -.1113341 .0614951 -1.810 0.076 -.2349784 .0123102 Igrad_6 | .058495 .0642929 0.910 0.367 -.0707746 .1877645 Igrad_7 | .1260635 .0961621 1.311 0.196 -.0672833 .3194104 Igrad_24 | .1188418 .085108 1. 396 0.169 -.0522793 .2899629 Igrad_30 | .1135478 .0596873 1.902 0.063 -.0064616 .2335571

172

Igrad_31 | -.0034473 .0822325 -0.042 0.967 -.1687868 .1618922 Igrad_32 | .2013207 .0679195 2.964 0.005 .0647594 .337882 a | -.0093058 .0564055 -0.165 0.870 -.1227165 .104105 b | -.2765195 .0607822 -4.549 0.000 -.3987302 -.1543088 c | -.1407444 .0599416 -2.348 0.023 -.261265 -.0202237 ER(a) | .4375952 .1420932 3.080 0.003 .1518977 .7232927 ER(d) | .358926 .1466186 2.448 0.018 .0641296 .6537224 _cons | 4.936801 .0862672 57.227 0.000 4.763349 5.110253 ------

173

Class VI

Source | SS df MS Number of obs = 239 ------+------F( 43, 195) = 44.24 Model | 4.50737029 43 .104822565 Prob > F = 0.0000 Residual | .462023491 195 .002369351 R-squared = 0.9070 ------+------Adj R-squared = 0.8865 Total | 4.96939378 238 .020879806 Root MSE = .04868

------lnfob2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ------+------mon_9 | .4808004 .0393318 12.224 0.000 .40323 .5583707 mon_10 | .3987872 .0378599 10.533 0.000 .3241198 .4734547 mon_11 | .2720498 .0363857 7.477 0.000 .2002897 .3438098 mon_12 | .28772 .0373111 7.711 0.000 .2141349 .3613052 mon_13 | .2803809 .0323982 8.654 0.000 .2164852 .3442767 mon_14 | .2526394 .0326889 7.729 0.000 .1881703 .3171084 mon_15 | .3219275 .0351447 9.160 0.000 .252615 .3912399 mon_16 | .3554651 .0346135 10.270 0.000 .2872003 .4237299 mon_17 | .3509778 .0322623 10.879 0.000 .28735 .4146056 mon_18 | .2560549 .0313599 8. 165 0.000 .1942067 .3179031 mon_19 | .1804365 .0319086 5.655 0.000 .1175063 .2433667 mon_20 | .253659 .0334865 7.575 0.000 .1876168 .3197011 mon_21 | .2773654 .0301621 9.196 0.000 .2178795 .3368512 mon_22 | .2561427 .0290914 8.805 0.000 .1987685 .313517 mon_23 | .2200397 .0246706 8.919 0.000 .1713843 .2686951 mon_24 | .1089774 .0336152 3.242 0.001 .0426814 .1752734 mon_25 | .0531726 .0286227 1.858 0.065 -.0032772 .1096223 mon_26 | .0734326 .0261167 2.812 0.005 .0219252 .12494 mon_27 | .079497 .0266933 2.978 0.003 .0268524 .1321416 mon_28 | .0737041 .0292773 2.517 0.013 .0159633 .1314449 mon_29 | -.0216039 .0328755 -0.657 0.512 -.0864411 .0432333 mon_30 | -.0676471 .0232885 -2.905 0.004 -.1135767 -.0217175 mon_31 | -.0318562 .0277488 -1.148 0.252 -.0865825 .0228701 mon_32 | -.1300586 .0259557 -5.011 0.000 -.1812486 -.0788686 mon_33 | -.1566827 .024775 -6.324 0.000 -.2055441 -.1078212 mon_34 | -.0297247 .0287219 -1.035 0.302 -.0863702 .0269208 mon_35 | -.0510785 .0214412 -2.382 0.018 -.093365 -.0087921 mon_36 | -.0293551 .0213926 -1.372 0.172 -.0715458 .0128356 mon_37 | -.0040053 .0263416 -0.152 0.879 -.0559563 .0479 457 mon_38 | -.0521065 .0266889 -1.952 0.052 -.1047425 .0005295 mon_39 | -.0300614 .0317726 -0.946 0.345 -.0927234 .0326005 mon_41 | .0164535 .0228665 0.720 0.473 -.028644 .061551

174

mon_42 | -.0104079 .0247818 -0.420 0.675 -.0592827 .0384668 mon_43 | .0146173 .0250404 0.584 0.560 -.0347675 .0640022 mon_45 | .0295568 .0244216 1.210 0.228 -.0186076 .0777212 Igrad_21 | .0529268 .0642478 0.824 0.411 -.073783 .1796365 Igrad_22 | .1005636 .0708161 1.420 0.157 -.0391002 .2402274 Igrad_50 | -.0503346 .0574699 -0.876 0.382 -.163677 .0630077 a | -.2339945 .0621065 -3.768 0.000 -.3564812 -.1115077 b | -.0559781 .0197339 -2.837 0.005 -.0948975 -.0170587 ER(b) | -.1552947 .066585 -2.332 0.021 -.2866139 -.0239754 Er(a) | 85.8054 17.84549 4.808 0.000 50.61046 121.0004 ER(c) | -.2528812 .0842174 -3.003 0.003 -.418975 -.0867874 _cons | 4.847293 .065252 74.286 0.000 4.718603 4.975984 ------

175

Class VII

Source | SS df MS Number of obs = 18 ------+------F( 10, 7) = 27.38 Model | .69146617 10 .069146617 Prob > F = 0.0001 Residual | .017679606 7 .002525658 R-squared = 0.9751 ------+------Adj R-squared = 0.9395 Total | .709145776 17 .041714457 Root MSE = .05026

------lnfob2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ------+------mon_10 | .911431 .4213873 2.163 0.067 -.0849915 1.907854 mon_12 | .9819512 .4368407 2.248 0.059 -.051013 2.014915 mon_21 | .66625 .1402224 4.751 0.002 .3346768 .9978232 mon_22 | .7863248 .1666638 4.718 0.002 .3922276 1.180422 mon_23 | .7328196 .1285666 5.700 0.001 .428808 1.036831 mon_26 | .4035901 .0577969 6.983 0.000 .2669221 .5402581 mon_32 | .2450916 .0893291 2.744 0.029 .0338619 .4563214 Igrad_1 | .0895154 .0315102 2.841 0.025 .0150056 .1640251 Igrad_3 | -.0705704 .0392203 -1.799 0.115 -.1633116 .0221708 ER(a) | -3.425629 2.559879 -1.338 0.223 -9.47878 2.627522 _cons | 4.446638 .33144 47 13.416 0.000 3.662896 5.23038 ------

176

APPENDIX 3 AUSTRALIAN WHEAT QUALITY CHARACTERISTICS

It is important to note that quality is a key influence in the demand for wheat as it has a crucial relationship to the end use product. Many demand studies have been conducted with relation to quality and class attributes such as those by Wilson (1994) and Ahmadi-Esfahani and Stanmore (1991 and 1994). Wilson (1994), in “Demand for Wheat Classes by Pacific Rim Countries”, highlights that Pacific Rim importers’ preferences have changed over time with a trend towards higher protein wheats (p 197).

Alternatively, Ahmadi and Stanmore (1994), in their paper “Quality Premiums for Australian Wheat in the Growing Asian Markets” notes some interesting results which suggest that there was a trend towards “a less quality rewarding world market” (p 247). This is a result of technological changes allowing for blending and adapting processing procedures to lower wheat qualities, regardless of the increasing incomes and changing consumer preferences (p 247).

“the end product from slightly lower quality wheat is not much worse than the higher quality wheat and (Asian markets) are therefore not willing to reward the higher quality wheat”

(Ahmadi-Esfahani and Stanmore, 1994, p 247).

INHERENT AND SEASONAL WHEAT QUALITY FACTORS Inherent or intrinsic quality characteristics are defined as those attributes which are hereditary and maybe controlled through genetic manipulation. Seasonal quality characteristics are those that change often, such as weather patterns or disease outbreaks, and cannot be easily controlled. Inherent characteristics, such as protein content and grain hardness maybe affected by seasonal conditions or other static factors, like soil fertility (Simmonds, 1989, p 31).

177

Table Inherent and Seasonal Quality Attributes of Wheat

Inherent Seasonal Protein type or quality Soundness and maturity Protein level or quantity Actual milling yield Grain hardness Actual protein content Milling yield Weather damage Resistance to weather damage Level of broken, shrivelled, frosted or green grains Seed coat cover Level of foreign seed Colour Presence of unmillable material Kernel size Presence of microrganisms (e.g. mould or insects) Water absorption Moisture content

(Source: Simmonds, 1989, p 31)

Uniformity of a shipment is also very important to end users, and strong preferences are seen for uniform shipments in the international market (Mercier, 1993 and Mercier and Hyberg, 1995). Uniformity can be best described as “a multidimensional attribute” (Smith, 2000, p 5), taking into account both inherent and seasonal characteristics. Uniformity generally holds across higher quality wheats and maybe thought of as a benchmark across wheat classes.

STANDARDS AND TESTS FOR INHERENT QUALITY ATTRIBUTES Milling quality is dependent on weight, shape, colour, vitreousity and hardness of the grain. Approximately 72-82% of the grain weight should be convertible into flour. Endosperm is the crucial factor within grain that produces flour “white flour is derived almost entirely from the endosperm of the grain” (Simmonds, 1989, p 33). Weight and shape of the grain indicates the level of endosperm, and hence millable material. These attributes are highly influenced by seasonal conditions such as temperature and often grains that are malformed

178 or shrivelled yield low levels of endosperm and hence result in poorer quality flour. Weight can be tested via hectolitre and 1000-kernel weight or chemical tests. Hecotlitre weight is dependent on “packing characteristics” and moisture content and is often less accurate though quicker and easier to perform than chemical procedures (Simmonds, 1989, p 34).

Colour is also a crucial determinant of milling quality. White and pale yellow tones are indicative of higher quality flours as opposed to yellow or brown pigments, which are usually a sign of high ash or bran content in the grain. Ash content is dependent on the nature of the “mineral matter” in the endosperm, and bran content is a function of the shape, size and density of the grain .Colour is also determined by the separability of the endosperm from the outerlayers of the kernel. Colour grade is measured using the Kent-Jones and Martin Flour Grader (Simmonds, 1989, p 40). Vitreousity or the translucentness of the wheat, is tested by “a subjective appraisal of the appearance of the grain when examined under a white lighted or black background” or by a cross sectional method (Simmonds, 1989, p 40). Vitreousness is important in relation to the colour of the wheat and are usually dependent on the hardness of the wheat, given beneficial growing conditions, though this may not always be the case.

Grain hardness plays a key role in the milling process as the brain layers need to be relatively easy to break down to ensure a maximum flour yield with appropriate colour results, a “clean separation of bran from endosperm” is achieved in harder wheats (Simmonds, 1989, p 34-35). Hard wheats usually have a more compact endosperm structure, which means that there is higher starch granule damage in the milling of hard wheats as opposed to soft wheat. Higher starch damage results in flour with better fermentation properties due to the increase sugar levels in the flour. Hardness is tested using the Particle Size Index (PSI) test or the Pearling Resistance test (Simmonds, 1989, p 39). PSI is the more common of these tests and works by measuring the amount of flour produced for the size of the particles. Larger particles are indicative of harder wheats due to the compactness of the endosperm.

179

Simmonds (1989) comments that “most Australian wheats already possess the desirable structural characteristics for good milling quality and improvement is continually being sought, and achieved, in newly released varieties” (p 35). It is also important to note that with recent technological advancement milling has become more efficient and technology has been developed to ensure that a maximum flour yield results from the milling of the majority of grain types. Rollers have been fashioned to ensure the release of endosperm from the remainder of the grain to produce good quality and coloured flours.

In the development of new varieties of grain, based on end use objectives and superior milling properties, the Buhler Laboratory Mill is used to test these new qualities (Simmonds, 1989, p 37). These small scale tests are conducted under controlled conditions by the Bread Research Institute (BRI), of Australia and research into matching breeding with end use characteristics is crucial for the continued development of Australia’s export markets.

AUSTRALIAN WHEAT CLASSIFICATIONS There are ten main wheat quality “receival standards” made by AWB Ltd., and these are measured by four basic quality criteria – protein levels, hardness, dough properties and milling qualities. It is important to note that protein is believed to be the most influential quality characteristic on end use products and often commands a premium (AWB, 1999a). Each of these ten classes are also divided into subclasses or grades depending on the specifity of the four basic quality criteria and the variety of wheat produced.

PRIME HARD Prime Hard is a top quality, reputable, high protein milling wheat, comprising of a selection of hard wheats with a protein level above 13% . Australian total production usually returns only 5% Prime Hard. End use products include superior Chinese style yellow alkaline noodles and Japanese Ramen noodles, making Prime Hard a perfect export to the Asian markets. The high protein levels also mean that Prime Hard can be used to produce some breads and may be blended with lower protein wheats for processing into other baked goods.

180

AUSTRALIAN HARD This wheat quality is also superior hard white wheat although the minimum protein leve l is lower at 11.5% . Over 14% of wheat produced in 1998/99 was Australian Hard standards. End uses include breads such as the European style pan and variety breads as well as Middle Eastern flat breads, Chinese steamed goods and alkaline noodles. This type of wheat is hence suitable for export to the Middle East and Egypt as well as Europe and Asia.

AUSTRALIAN PREMIUM WHITE (APW) Similar again to the Prime Hard and Australian Hard, APW is described as “a unique blend of hard grained wheat varieties selected to ensure consistently high milling performance and flour quality” (AWB, 1999a). The protein levels of this wheat begin at 10% making it suitable for manufacture of fresh and dried Asian noodles, including the Hokkien noodle, as well as Indian and Middle Eastern flat breads and Chinese steamed products. 31% of the 1998/99 wheat crop was suitable for APW classification.

AUSTRALIAN STANDARD WHITE (ASW) ASW is the usually the broadest class of Australian wheat. The 1998/99 crop consisted of 32% ASW. Having a lower protein level this wheat is reputed for its versatility and high capital returns. End uses include flat breads eaten in the Middle East and India, as well as European style breads and Chinese steamed breads.

NOODLE WHEAT Noodle Wheat is separated by the AWB for the manufacture of Japanese white salted noodles, such as the Ramen and Udon noodles. These noodles require soft grained wheats, mainly produced in Western Australia and certain areas in Victoria, specifically grown for the Japanese and South Korean market. Australian Prime Hard from NSW and Queensland are also further classified into ‘Chinese Noodle’ wheat, again primarily for the Japanese market (AWB, 1999a).

181

AUSTRALIAN SOFT WHEAT These wheats have a lower protein level (maximum of 9.5%) (AWB, 1999a) and are suitable for the production of confectionery and baked goods such as biscuits, pastries and snack foods.

DURUM WHEAT Durum wheat is a selection of wheats with “vitreous, amber coloured kernels” (AWB, 1999a). Durum wheat is divided into 3 different categories dependent on protein levels, No. 1 has a minimum of 13% protein, No. 2, 11.5% and No. 3, 10%. The major end use of Durum wheat includes fresh and dried pasta products due to the wheat’s characteristic amber or yellow coloured pigment and the high water absorption. Durum wheat is primarily produced in NSW and South Australia and sold mainly in the domestic market, although recently their has been potential for development of an export market.

AUSTRALIAN GENERAL PURPOSE (GP1) GP1 consists of any wheat unable to fit the above quality classification due to higher screenings, lower falling number, lower kernel or test weights , fungal or weather (e.g. frost) damage or the presence of a large amount of unmillable material (e.g. weeds). GP1 usually consists of a low percentage of the Australian crop and is commonly used for blending especially if protein levels are high. The end uses of GP1 are as baked goods after blending with higher quality wheats. The advantage of blending often means a cost reduction to the processor.

PH5 (HIGH SCREENINGS) This is wheat which has very high protein levels however, is not able to meet the Prime Hard standard as a result of high screening levels. PH5 wheat consists of slightly reduced test weights and greatly reduced kernel weight which leads to a decrease in milling quality although flour extraction is only slightly less (2% less) than that of Prime Hard wheat (AWB, 1999a). The benefits of PH5 are the very high protein levels which produce very good quality end products similar to that of Prime Hard as well as providing a cost advantage to the processor.

182

AUSTRALIAN FEED Wheat is often used in the production of animal feeds. Feed wheat requires a high protein and gluten levels and well balanced nutrient content and is often manufactured into pellets and blended with other wheat (e.g. GP1) and non-wheat products (AWB, 1999a).

Table Differentiation of Wheat Class Produced per State as a Percentage of the Total Wheat Produced for that Class, 1997-98

Prim e Hard Hard

APW

ASW

(Source: ABARE, 1999, p 219)

183

WHEAT QUALITY REQUIREMENTS FOR SPECIFIC END USES

MILLING QUALITY Prior to human consumption wheat must be milled and converted to flour for further processing. Milling quality refers to the characteristics of the wheat required in order to produce different quality flours for the manufacture of different end use products. These attributes include flour yield from a given grain weight, rate at which milling can occur whilst achieving maximum yield, the colour of the flour and the moisture level of the grain prior to milling (Simmonds, 1989, p 22).

There are also four important dough49 properties crucial to the production of end use goods:

1. Protein Strength; 2. Water Absorption; 3. Flour Colour, and 4. Fermentation Properties.

(Simmonds, 1989, p 22-23)

Protein strength refers to the physical qualities of the dough and is directly related to the protein and gluten levels in the wheat used. Water absorption is the amount of water absorbed in the dough making process and is associated with the level of protein, the amount of starch and the levels of non-starch carbohydrates or pentosans (Simmonds, 1989, p 23). The optimal colour for flour is white or slightly yellow. Flours which are darker in colour usually have high levels of bran present and these darker coloured flours often lead to the presence of undesirable characteristics in end use goods, such as low loaf volumes (Simmonds, 1989, p 23). Fermentation properties relate to the sugar levels (either added

49 Dough is a mixture of flour and water and is a direct function of the strength and elasticity of gluten (or hydrated proteins) in wheat.

184 sugar or hydrolised granules of damaged wheat starch) which impacts on the production of breads (leavened or unleavened).

NOODLE MANUFACTURING Noodles are a crucial component of the Chinese, Japanese and South East Asian diet and wheat noodles are common. Wheat noodles are made from wheat flour, water and salt and these three basic ingredients in varying quantities and using different varieties/qualities of wheat flour are used in the manufacture of several different types of noodles. Australian wheat is popular in the noodle manufacture of noodles in Asia for several reasons including proximity of supply, availability of desired quality charactersitics (colour (white), good milling properties, low moisture content, required level of starch damage and superior dough strength (Simmonds, 1989, p 26). The Asian consumer is said to prefer a “high eating quality and colour, which must be a clear pale yellow, free from any discolouration” (Simmonds, 1989, p 26).

The wheat protein level is a crucial characteristic in noodle making, with an optimum level required for each different noodle type. Protein is crucial to the flexibility and firmness in the ready to eat product. Chinese noodles (including Hokkien style and wet or dry noodles), require medium to hard wheats with a protein level of 11% (typically Queensland and northern New South Wales Prime Hard varieties), which produce a good coloured flour and an elastic dough (Simmonds, 1989, p 27). Japanese or White Noodles are prepared from lower protein level medium wheats (8-10.5%), producing a weaker flour but must be made from a low colour grade with minimal ash or bran content (Simmonds, 1989, p 28). Instant noodle, being steamed and fried, as opposed to be fresh or dried, are manufactured from softer wheats than other noodles, for example Australian Soft or Australian Standard White wheats from Victoria, NSW or Western Australia. Softer wheats generally have a lower protein level and have lower water absorption and starch damage properties.

185

The noodle making properties of flour are tested by research bodies such as the Australian BRI. Tests include the creation of noodle dough from flour with a pH of 9.5 and a moisture content around 35% (Simmonds, 1989, p 50). The dough is then sheeted into 1.5mm thick sheets and the colour examined with a Hunterlab Colour Difference Meter to ensure a brightness of greater than or equal to 70 and yellowness ranked between 45 and 50 (Simmonds, 1989, p 50). Eating quality is generally associated with amylose (starch), content and possess smooth, soft and firm dough characteristics (Simmonds, 1989, p 50).

186

REFERENCES

Australian Bureau of Agricultural and Resource Economics (ABARE), 1999, Australian Commodity Statistics 1999, Canberra

ABARE, 2000, Australian Commodity Statistics 2000, Canberra

ABARE, 2001, Australian Commodity Statistics 2001, Canberra

Ackerman, K.Z. and Dixit, P.M., 1999, “An Introduction to State Trading in Agriculture”, Economic Research Services, United States Department of Agriculture, AER-783, USA

Agriculture and Agri-food Canada (AAFC), 1998, “The Australian Wheat Board”, Bi-weekly Bulletin, 11(2), 30th January, 1998, Canada

Ahamdi-Esfahani, F.Z. and Stanmore, R.G., 1992, “Is Wheat a Homogeneous Product? A Comment”, Canadian Journal of Agricultural Economics, 40(1), pp141-146

Ahmadi-Esfahani, F.Z., 1994, “Wheat Marketing and Trade: Further Issues for Research – A Comment on Ryan”, Review of Marketing and Agricultural Economics, 62 (1)

Alouze, C.M., Watson, A.S. and Sturgess, N.H., 1978, “Oligopoly Pricing in the World Wheat Market”, American Journal of Agricultural Economics, 60, pp173-185

Allen Consulting Group (ACG), 2000a, The Wheat Marketing Act 1989: The Economic Impact of Competitive Restrictions, Report to the Independent National Competition Policy Review Committee, Sydney.

187

Allen Consulting Group (ACG), 2000b, The Wheat Marketing Act 1989: The Social Impact of Competitive Restrictions, Report to the Independent National Competition Policy Review Committee, Sydney.

Alston, J.M. and Gray, R., 2000, “State Trading versus Export Subsidies: The Case of Canadian Wheat”, Journal of Agricultural and Resource Economics, 25(1), pp 51-67

Amey, E.B., 1998, “Gold”, Metal Prices in the United States Throughout 1998, United States Geological Society, July 1998, Gold pp 49-53

Anania, G., Bohman, M. and Carter, C.A., 1992, “United States Export Subsidies in Wheat: Strategic Trade Policy or Expensive Beggar-Thy-Neighbor Tactic?”, American Journal of Agricultural Economics, 74(3), pp 534-545

Anderson, J., 1997, Department of Primary Industries and Energy Press Release, 97/37A, 17 April 1997

Antle, J.M. and Smith, V.H. (eds.), 1999, The Economics of World Wheat Markets, CABI Publishing, USA

Australian Financial Review, 2001, AWB: Time to Sow Right Seed, 17th January, 2001, www.afr.com.au

Austroads, 2000, “Roadfacts 2000”, Austroads Incorporated, Publication Number AP- G18/00

Australian Wheat Board (AWB) Ltd., 1999, “The Australian Grains Industry – An Introduction”, www.awb.com.au

188

AWB Ltd., 2001a, “Community Education – Transporting Grain”, www.awb.com.au

AWB Ltd., 2001b, “AWB National Pool Performance Report 2000-01”, www.awb.com.au

AWB (International) Limited (AWB(I)), 2000, Review of the Wheat Marketing Act 1989, Submission to the NCP Review, Melbourne.

Baldwin, R.E., 1970, “Non-Tariff Distortions of International Trade”, Brookings Institue, USA

Baum, C.F., Caglayan, M. and Ozkan, N., 2003, “The Impact of Macroeconomic Uncertainty on Trade Credit for Non-Financial Firms”, Department of Economics, Boston College, Working Paper, http://fmwww.bc.edu/ec-p/wp566.pdf.

Beard, R. and Purcell, T., 1996, “An Institutional Analysis of Partial Deregulation of the Australian Wheat Industry under Bertrand Competition”, Agricultural Economics Discussion Paper 2/96, University of Queensland

Bishop, M.; Kay, J.; and Mayer, C. (eds.), 1994, Privatisation and Economic Performance, Oxford University Press, Oxford, UK

Booz, Allen and Hamilton, 1995, Milling Wheat Project, Grains Council of Australia’s National Grain Marketing Strategic Planning Unit, January, 1995 Australia

Bos, D.,1986, Public Enterprise Economics: Theory and Application, North-Holland, Oxford

Bos, D., 1991, Privatization: A Theoretical Treatment, Clarendon Press, Oxford

189

Brenner, R., 1987, “State Owned Enterprises”, in Rivalry: In Business, Science among Nations, Cambridge University Press, Sydney, Australia

Burton, M.P. and Lobb, A.E., 2000, “2000 National Competition Policy Review of the 1989 Wheat Marketing Act -Price Premiums and the AWB (International) Ltd.”, Working Paper, University of Western Australia, Perth; from Burton, M.P. (with Lobb, A.E.), 2000, “Independent Review Report”, for the Independent Review Committee for the 2000 National Competition Policy Review of the Wheat Marketing Act (1989).

Carter, C. & Schmitz, A., 1979, “Import Tariffs and Price Formation in the World Wheat Market”, American Journal of Agricultural Economics, 61, August 1979

Carter, C.A., Loyns, R.M.A. and Ahmadi-Esfahani, F., 1986, “Varietal Licensing Standards and Canadian Wheat Exports”, Canadian Journal of Agricultural Economics, 34, November, 1986, Canada

Carter, C. A., 1993, “An Economic Analysis of a Single North American Barley Market”, Canadian Journal of Agricultural Economics, 41(3),Canada

Carter, C. A. and Loyns, R. M. A., 1996, The Economics of Single-Desk Selling of Western Canadian Grain, Food and Rural Development, Alberta Agriculture, Edmonton.

Carter. C.A.; Loyns, R.M.A. and Berwald, D., 1998, “Domestic Costs of Statuatory Marketing Authorities”, American Journal of Agricultural Economics, 80, May 1998, pp 313-324

Carter, C., MacLaren, D. and Yilmaz, A., 1999, “How Competitive is the World Wheat Market?”, Working Paper, University of California, Davis

190

Carter, C.A. and Wilson, W.W., 1999, “Emerging Differences in State Grain Trading”, in Antle, J.M. and Smith, V.H. (eds.), 1999, The Economics of World Wheat Markets, CABI Publishing, USA, pp 203-220

Centre for International Economics (CIE), 1997, “Review of the Victorian and South Australian Barley Marketing Act 1993”, Final Report prepared for the Department of Natural Resources and Environment, Victoria, and Primary Industries South Australia

Clark, D., 1995, “Microeconomic Reform” in Kriesler, P (ed) (1995), The Australian Economy: The Essential Guide, Allen and Unwin, Sydney

CRU International Ltd., 1996, “Gold”, CRU International Ltd Quarterly Market Service, November 1996

Canadian Wheat Board (CWB), www.cwb.ca

CWB, 2003, “The Canadian Wheat Board 2002-03 Statistical Tables”, http://www.cwb.ca/en/publications/students_researchers/pdf/2002- 03_full_english_statistics.pdf

Department of Transport and Regional Services (DoTRS), 2000, “The Commonwealth’s Transport Directions - Task and Outlook”, Commonwealth of Australia.

Dixit, P.M. and Josling, T., 1997, State Trading in Agriculture: an Analytical Framework, International Agricultural Trade Research Consortium on Working Paper 4, July 1997.

Economic Planning and Advisory Committee (EPAC), 1989, “Transport and Australian Industry” EPAC Report, June 1989

191

Economic Research Services (ERS), 1998, “Agriculture in the WTO”, USDA, WRS-98-44, December 1998, USA

Foreign Agricultural Services (FAS), 1998a, “US Grain Exports to Malaysia Suffer during Currency Crisis”, FAS Grains Circular, March 1998, www.fas.usda.gov/grain/circular/1998/98-03/

FAS, 1998b, “China Seeks Balance in Providing for its Expanding Grain Needs” FAS Grains Circular, April 1998, www.fas.usda.gov/grain/circular/1998/98-04/

FAS, 2000,

Fraser, R.W., 1989, “Uncertainty and the Positive Theory of Public Enterprise”, Bulletin of Economic Research, 41(2), pp147-155

Fraser, R., 1991, “Privatisation in the United Kingdom: Lessons for Australia?”, Economic Papers, September 1991, 10(3), pp30-37

Fraser, R., 1994a, “Privatisation: Price-capping and Reliability”, Utilities Policy, April 1994, 4(2), pp121-127

Fraser, R., 1994b, “Price, Quality and Regulation: An Analysis of Price-Capping and the Reliability of Electricity Supply, Energy Economics, July 1994, 16(3), pp 175-183

Fraser, R., 1996, “Privatisation in the United Kingdom: Lessons for Australia 2”, Economic Papers, December 1996, 15(4), pp14-19

Furtan, W.H., 1995, “Agriculture and Federation”, Canadian Journal of Agricultural Economics, 43(4), December, pp 545-550.

192

FXHistory: Historical Currency Exchange Rates (http://www.oanda.com/convert/fxhistory)

Gold Fields Mineral Services Ltd., 1997, “Gold 1997”, Gold Fields Mineral Services Limited, London

Goldberg, P.K. and Knetter, M.M., 1999, “Measuring the Intensity of Competition in Export Markets”, Journal of International Economics, 47(1), February, pp 27-60

Gropp, L., Hallam, T., and Manion, V., 2000, “Single Desk Marketing: Assessing the Economic Arrangements”, Productivity Commission Staff Research Paper, July 2000, www.pc.gov.au

Govindan, A., 2000, “India – Grain and Feed April Update 2000”, GAIN, FAS, Report #IN0016, USA

Gujarati, D.N., (3rd ed), 2002, Basic Econometrics, McGraw Hill, London

Hanson, S.D. and Ladd, G.W., 1991, “Robustness of the Mean-Variance Model with Truncated Probability Distribution”, American Journal of Agricultural Economics, May 1991, 73(2), pp436-445

Hicks, P. and Ireland, I., 1997, Wheat Marketing Act Amendment Bill 1997, Bills Digest No. 43, Bills Digest Service, Information and Research Services, Department of the Parliamentary Library, Commonwealth of Australia

House, M., 2000, “Thailand – Grain and Feed Annual, 2000”, GAIN, FAS, Report #TH0026, USA

193

Howard, J., 1997, “Investing for Growth”, Address by the Prime Minister, The Hon John Howard MP, National Press Club Canberra, 7th December, 1997 (http://www.isr.gov.au/growth/html/speech.html).

Industry Assistance Commissio n (IAC), 1977, “Wheat Stabilisation”, IAC Report, June 1977

IAC, 1978, “Wheat Stabilisation”, IAC Report, July 1978

IAC, 1984, “The Wheat Industry”, IAC Report, July 1984

IAC, 1988, “The Wheat Industry” Report # 411, February 1988

IAC, 1988a, “The Wheat Industry – Conclusions and Recommendations”, IAC Report, February, 1988

IAC, 1988b, “The Wheat Industry”, IAC Report, April 1988

IAC, 1989, “Public Rail Services – Inquiry into Government Charges, IAC Report, February 1989

Independent Committee (IC), 1993, “Port Authority Services and Activities”, IC Inquiry Report No. 31, AusInfo, Canberra, May 1993

Independent Review Committee (IRC), 2000, “National Competition Policy Review of the Wheat Marketing Act 1989”, NCP - WMA Review Committee, December 2000, Canberra

Industries Commission (IC), 1991, “Rail Transport Report”, Report No. 13, August 1991

194

International Monetary Fund (IMF), 1999, “World Economic Outlook”, October 1999

Ireland, I., 1998, “Wheat Marketing Legislation Amendment Bill 1998” Bills Digest #220, 1997-1998, Information and Research Services, www.aph.gov.au/library/pubs/bd/1997- 98/98bd220.htm

Irving, M., Arney, J and Lindner, B., 2000, “National Competition Policy Review of the Wheat Marketing Act, 1989”, NCP – WMA Review Committee, for the Department of Agriculture, Fisheries and Forestry, Canberra

Johnson, D.M. and Niniek, S.A., 2000, “Indonesia – Grain and Feed Annual Report 2000”, GAIN, FAS, Report # ID0016, USA

Johnston, J.H., 1984, “Forum: Wheat Marketing Symposium 1984, An Introduction” and “An Epilogue”, Review of Marketing and Agricultural Economics, 52(2), pp 101-105 and pp 129-135;

Joint Industry Submission Group (JISG), 2000, Australian Wheat: It’s Time for Choice, A submission to the Independent Review Committee reviewing the Wheat Marketing Act 1989 under National Competition Policy, Centre for International Economics, Canberra.

Just, R.E., Schmitz, A. and Zilberman, D., 1979, “Price Controls and Optimal Export Policies Under Alternative Market Structures”, American Economic Review, 69, pp 706-714.

Karp, L.S. and McCalla, A.F. 1983 “Dynamic Games and International Trade: An Application to the World Corn Market”, American Journal of Agricultural Economics, November 1983

195

Katz, M., 1984, “Price Discrimination and Monopolisitc Competition”, Econometrica, 52(6), November, pp 1453-1472

Knetter, M. M., 1989, “Price Discrimination by U. S. and German Exporters”, American Economic Review 79, pp. 198-210.

Kolstad, C.D. and Burris, A.E., 1986, “Imperfectly Competitive Equilibria in International Commodity Markets”, American Journal of Agricultural Economics, 68(1), February, pp27- 36

Kostecki, M.M. (ed.), 1982, State Trading in International Markets, St. Martin’s Press, USA

Kraft, D.F., Furtan, W.H. and Tyrchniewicz, E.W., 1996, “Performance Evaluation of the Canadian Wheat Board”, Technical Report, Canadian Wheat Board, Canada

Kronos Corporate, 2002, A Review of Structural Issues in the Australian Grain Market, www.capgrains.com.au/grain/kronos_report/ full_kronos_report.pdf

Larue, B. and Lapan, H.E., 1992, “Market Structure, Quality and World Wheat Market”, Canadian Journal of Agricultural Economics, 40, pp311-328

Love, H.A. and Murniningtyas, E., 1992, “Measuring the Degree of Market Power Exerted by Government Trade Agencies”, American Journal of Agricultural Economics, 64, August 1992, USA

Lucas, J.M., 1983, “Gold”, The Minerals Yearbook, Vol 1, US Bureau of Mines, pp 369-396

Lucas, J.M., 1984, “Gold”, The Minerals Yearbook,1, US Bureau of Mines, pp 385-441

196

Lucas, J.M., 1985, “Gold”, The Minerals Yearbook,1, US Bureau of Mines, pp 405-435

Lucas, J.M., 1988, “Gold”, Mineral Commodity Summaries 1987, US Bureau of Mines, pp 421-458

MacAulay, T.G., 2000, “Competition Policy in Agriculture: A Review of Methods”, contributed paper presented the 44th Australian Agricultural and Resource Economic Society Conference, 23rd-25th January, 2000, University of Sydney

McCalla, A.F., 1966, “A Duopoly Model of World Wheat Pricing”, Journal of Farm Economics, 48(3)

McCalla, A.F. and Schmitz, A., 1979, “Grain Marketing Systems: The Case of the United States versus Canada”, American Journal of Agricultural Economics, 61, May 1979

McCorriston, S. and MacLaren, D., 2002, “Perspectives on the State Trading Issue in the WTO Negotiations”, European Review of Agricultural Economics, 29(1), pp 131-145.

Meyers Strategy Group (MSG), 1996, “Economic Analysis of the Value of the Single Desk”, Report to the Australian Barley Board, Sydney

Miller, G.F. and White, G., 1980, “The Seventh Wheat Industry Stablisation Scheme, Evolution and Economic Effects”, Paper contributed to the 24th Annual Conference of the Australian Agricultural Economics Society Conference, Adelaide

National Competition Policy (NCP), 2001, “Agriculture and Related Activities”, Ch 13, NCP Assessment, Commonwealth of Australia

197

NCP Review of the WMA Secretariat, 2000, “Wheat Review Committee Delivers Final Report to Minister”, Media Release, Department of Agriculture, Fisheries and Forestry Australia, 22 December, 2000, www.affa.gov.au

Organisation for Economic Co-operation and Development (OECD), 2001, Agricultural Policies in OECD Countries: Monitoring and Evaluation, Paris

Paarlberg, P.C. and Abbott, P., 1986, “Oligopolistic Behavior by Public Agencies in International Trade: The World Wheat Market”, American Journal of Agricultural Economics, 68(3), August, pp 528-543.

Pick, D.H. and Park, T.A., 1991, “The Competitive Structure of US Agricultural Exports”, American Journal of Agricultural Economics, February, 1991

Piggott, R.R., 1992, “Some Old Truths Revisited”, Australian Journal of Agricultural Economics, 36, p 117-140

Piggott, R. and Edwards, G., 2000, Methodology to be Employed in Evaluating Economic Impacts of the WMA, Report submitted to the National Competition Policy Review of the Wheat Marketing Act Committee, Canberra.

Productivity Commission (PC), 1998, “International Benchmarking of the Australian Waterfront”, International Benchmarking Report, AusInfo, Canberra

PC, 2000, “Progress in Rail Reform”, AusInfo, Canberra

PC, 2002a, “Trends in Australian Infrastructure Prices, 1990-91 to 2000-01”, Performance Monitoring, AusInfo, Canberra, May 2002

198

PC, 2002b, “Economic Regulation of Harbour Towage and Related Services”, Position Paper, Canberra, June 2002

Quiggin, J., 1996, Great Expectations: Microeconomic Reform in Australia, Allen and Unwin, Sydney

Rees, R., 1984, “A Positive Theory of the Public Enterprise”, in Marchand, M., Pestieau, P. and Tulkens, H. (eds), 1984, The Performance of Public Enterprises, North-Holland, Amsterdam

Roberts, I. and Doyle, S., 1996, “US Farm Legislation: US Federal Agricultural Improvement and Reform Act of 1996”, Australian Commodities Forecasts and Issues, 3(2), June, pp 210-24

Robinson, J., 1933, The Economics of Imperfect Competition, Macmillan and Co., London

Robinson, J., 1950, The Economics of Imperfect Competition, Macmillan and Co., London (pp 20-21)

Roskill Information Services Ltd., 1995, “Gold – Market Update Analysis and Outlook”, Roskill Information Services Ltd.

Rozelle, S.D. and Huang, J., 1999, “Wheat in China: Supply, Demand and Trade in the 21st Century”, in Antle, J.M. and Smith, V.H. (eds.), 1999, The Economics of World Wheat Markets, CABI Publishing, USA, pp 145-173

Ryan, T.J., 1984, “Wheat Marketing”, Review of Marketing and Agricultural Economics, 52(2), pp 117-127

199

Ryan, T.J., 1994, “Marketing Australia’s Crop: The Way Ahead”, Review of Marketing and Agricultural Economics, 62, pp107-121

Salop, S. and Stiglitz, J., 1977, “Bargains and Ripoffs: A Model of Monopolistically Competitive Price Dispersion”, Review of Economic Studies, 44, pp 493-510

Sarris, A. and Freebairn, J., 1983, “Endogenous Price Policies and International Wheat Prices”, American Journal of Agricultural Economics, 65(2), pp 214-224

Schmalensee, R., 1981, “Output and Welfare Implications of Monopolistic Third-Degree Price Discrimination”, American Economic Review, 71, March, pp 242-247

Schmitz, A., Gray, R., Schmitz, T.G. and Storey, G., 1997, “The CWB and Barley Marketing: Price Pooling and Single Desk Selling”, Technical Report, Canadian Wheat Board, Canada

Schmitz, T.G., and Gray, R., 2000, “State Trading Enterprises and Revenue Gains from Market Power: The Case of Barley Marketing and the Canadian Wheat Board”, Journal of Agricultural and Resource Economics, 25(2), pp 596 –615 (p 611).

Shipping Australia Ltd., 2002, “Submission to the Productivity Commission by Shipping Australia Ltd., Regarding the Inquiry into Economic Regulation of Harbour Towage and Related Services in Australia”, Attachment C; in PC, 2002b, “Economic Regulation of Harbour Towage and Related Services”, Position Paper, Canberra, June 2002

Simmonds, D.H. (1989), Wheat and Wheat Quality in Australia, CSIRO, Australia

200

Skousen, M. (1997), “Which is the Best Inflation Indicator: Gold, Oil or the Commodity Spot Index?” The Freeman, 47(2) The Foundation for Economic Education, Inc., February 1997.

Smith, V.H., 2000, “Wheat Quality and Wheat Yeilds: Trade-Offs among Price, Yield, Profit and Risk”, Special Report no. 5, Trade Research Center, Montana State University, USA

Stout, J. and Alber, D., 2003, “ERS/Penn State WTO Model Documentation”, Draft 9, Working Paper, USDA, ERS, http://trade.aers.psu.edu/index.cfm

Tarchalski, K. et al., 1996, “ASEAN Wheat Markets: Prospects for Their Liberalisation”, Australian Commodities Forecasts and Issues, 3(1), March, pp63-80

The Hon. Warren Truss, MP, 2000 in “Wheat Marketing Review Recommends Retaining the Single Desk Arrangements”, Media Release (AFFA00/WT), Department of Agriculture, Fisheries and Forestry Australia, 22 December, 2000, www.affa.gov.au

Turner, S. et al., 2000, “Grains – Outlook to 2004-05”, Australian Commodities Forecasts and Issues, 7(1), March, pp30-44

Ulrich, A., Furtan, W.H. and Schmitz, A., 1987, “The Cost of a Licensing System Regulation: An Example from Canadian Prairie Agriculture”, Journal of Political Economy, 95(1), USA

Vanzetti, D., 1991, “Policy Coordination in the World Wheat Market”, Agricultural Economics Discussion Paper, 4/91, University of Western Australia, Perth

201

Varian, H., 1985, “Price Discrimination and Social Welfare”, The American Economic Review, 75(4), September, pp 870-875

Vickers, J. and Yarrow, G., 1988, Privatization – An Economic Analysis, The MIT Press, Cambridge, MA, USA

Vuong, Q.H., 1989, “Likelihood Ratio Tests for Model Selection and Non-nested Hypotheses”, Econometrica, 57(2), March, pp307-33

Wade, J. and Zhang, J., 2000, “China – Grain and Feed March Update, 2000”, Global Agricultural Information Network (GAIN), FAS, Report #CH0015, USA

Wait, M and Ahmadi-Esfahani, F., 1996, “How has the Domestic Wheat Market Changed since Deregulation?”, Review of Marketing and Agricultural Economics, 64(3), December 19996, pp 319-324

Water Industry Reform Authority (WIRA), 1992, “The Waterfront”, Canberra

Watson, A.S., 1984, “Wheat in 1984”, Review of Marketing and Agricultural Economics, 52(2), pp 107-115.

Watson, A.S., 1999, “Grain Marketing and National Competition Policy: reform or reaction?”, Australian Journal of Agricultural and Resource Economics, 43(4), pp 429-455, December, 1999.

Wilson, W.W., 1989, “Differentiation and Implicit Prices in Export Wheat Markets”, Western Journal of Agricultural Economics, 14(1)

202

Wilson, W.W., 1994, “Demand for Wheat Classes by Pacific Rim Countries”, Journal of Agricultural and Resource Economics, 19(1)

Wilson, W.W. and Dahl, B.L., 1998, “Grain Quality and North American Hard Wheat Exports”, Research Discussion Paper No. 15, September 1998, North Dakota State University, USA

Wilson, W.W., Dahl, B.L. and Johnson, D.D., 2000, “Transparency and Bidding Competition in International Wheat Trade”, Working Paper, North Dakota State University, USA

World Trade Organisation (WTO), 1994, The Results of the Uruguay Round of Multilateral Trade Negotiations: The Legal Texts, Geneva.

WTO, 1995, “Operations of State Trading Enterprises as they Relate to International Trade”, Background paper to the Secretariat, G/STR/2, October 26th 1995, Switzerland

203