This research was supported by generous financial assistance from the Competition Commission South Africa in the form of a student loan and the Barcelona Graduate School of Economics in the form of the Barcelona Graduate School of Economics visiting program. Part of the research reflected in this thesis was carried out as a visiting graduate student at the Barcelona Graduate School of Economics, and I am grateful to Massimo Motta for making this arrangement possible. All these years of research would not have been so enjoyable without the friendship and company of many people at the Competition Commission South Africa and the University of KwaZulu-Natal.
This is a collection of related essays on collusion following various competition policy interventions in the recent past. I ask whether observed prices and quantities in the flour industry reflect the switch from collusive to non-collusive behavior and test this empirically. Essay three examines the strategic behavior adopted by the cartel and in particular efforts by Pioneer Foods to create artificial barriers to entry.
Essay five evaluates this claim by examining the design and effectiveness of the rebate tool from a comparative perspective. To enhance our understanding of competition policy and its administration, economic understanding will be combined with an appreciation of case-specific issues and the South African legal framework.
INTRODUCTION
- The law on collusion in South Africa
- The economics of collusion
- Detecting collusion
- Corporate leniency policies
- Industry background
- Introduction to the essays
In the EU, the total fine does not exceed 10 percent of the company's total turnover or worldwide sales. In the short run, deviating becomes less profitable because it is more difficult to attract customers from the competing firm by lowering the price. Before the adoption of CLP in February 2004, few cartels had been investigated and prosecuted under the Competition Act.
As a result, the CLP creates a race to the door by rewarding the first member of the cartel to produce evidence of the cartel. CLP applies to any cartel conduct that falls within the scope of the Act. As a prelude to the essays, it is useful to begin with a brief overview of the grain industry.
With the establishment of the cartel, incentives were immediately created for each firm to cheat. With repeated interaction, the ability to enforce collusive agreements became an important part of the cartel arrangements.
ESSAY ONE
THE SOUTH AFRICAN WHEAT FLOUR CARTEL: OVERCHARGES AT THE MILL 17
The Wheat Board was the main intermediary between the farm gate and the processing level for wheat products. Marketing of wheat was regulated through a single channel marketing system administered by the Wheat Board. But instead of competing, the mills simply replaced state regulation with private regulation.34 The milk cartel was exposed in 2007.
The cartel was exposed when Premier Foods, one of the firms involved in the cartel, applied for and was granted business reduction under the CLP. The cartel fixed the price of flour, bread and cornmeal and allocated customers in flour and bread from 1999 to 2007. Before analyzing the effect of the wheat flour cartel on prices, an understanding of both the wheat flour market and the cartel system is an essential prerequisite.
After the cartel's liability is established, the task is to determine overprice estimates due to price fixing using a reduced-form price equation. In particular, I find that the overcharges for independent bakeries in flour range from 7 per cent. to around 42 per cent.
The industry
With the outbreak of the Second World War, the Wheat Board introduced subsidies for Grade A wheat to increase production. The Wheat Board ceased to regulate the prices and marketing of products derived from winter grain processing in 1991, although it continued to regulate the production price of wheat. Specifically for wheat, the Wheat Board was the main intermediary between the farm gate and the processing level for wheat products.
The changes included the closure of the Wheat Board, the shift from quantitative trade restrictions to tariffs, and the gradual reduction of tariffs. Wheat represents one of the most important field crops in South Africa.44 The wheat industry is important because it contributes significantly to the total gross value of agricultural production. The wheat milling process consists of three stages, each with its own specific type of machinery.
In the breaking process, roller mills are used to gently break open the wheat kernel to prepare the wheat for further processing, with the aim of removing as much of the bran as possible from the endosperm. The main products produced from the wheat milling process in South Africa include brown and white bread flour (used for baking bread, white bread flour accounts for about 40 percent of sales, while brown bread flour accounts for about 26 percent of sales), industrial flour ( sold only to industrial users and not available to end customers, it accounts for about 1 percent of flour sales), wheat waste (used for grain and/or sold to the feed industry) and cake flour (used for baking cakes and accounts for about 30 percent of sales).
The flour cartel
Pursuant to the consent agreement, the Court fined Tiger Brands for its role in the bread cartel. Tiger Brands admitted its role in the bread cartel and presented further evidence about the paint cartel. 54 The court fined Tiger Brands R98 million for its role in the bread cartel.
Competition in the bread market can be characterized as involving a cartel (given that the cartel controls more than 98% of the market, this is akin to a monopoly baker) with a competitive edge. The cartel in the flour market harms the independent bakeries (who demand flour to produce bread), industrial flour buyers (industrial flour is sold to industrial users in bulk and is not available to end customers) as well as indirect buyers of flour (consumers who buy bread for their Spending). In the absence of a pass-through effect, the harm caused by a cartel to direct customers is the sum of areas a, b and c.
Bread consumers pay a price 𝑝1 when they would have paid 𝑝0 in the absence of the cartel. The analysis below assumes that there is no pass-through effect in the independent bakery market.
Estimating wheat flour price overcharges to independent bakeries
A comparison of prices in the same market during and after the cartel; and (3) both unaffected pre- and post-cartel periods (comparison "before, during, and after" the cartel). The regression model is then used to predict but-for prices during the cartel period. This is because the members of the cartel may be able to tacitly enter into agreements after the cartel ceases to exist.
The private data is from the National Chamber of Mills ("NCM") and companies involved in the cartel. It is hard to tell from the numbers the likely influence of the cartel. 79 Suppose we want to compare the effect of the cost of wheat on the price of flour between the cartel period (𝐷𝑖) and the no-cartel period (𝐹𝑖).
To illustrate using the results for brown bread flour in the Western Cape province (column 1 of Tables 1-6), the long-run effect of the cartel on the equilibrium price is approximately 27.6 percent. While for Gauteng province cake flour, the cartel's long-term effect on the equilibrium price is approximately 13 percent.
Conclusion
Best practice suggests that whenever used, one should assess the results of each approach in light of its limitations, perform a sensitivity analysis to assess the robustness of the results, and reconcile with more direct evidence that are available (for example internal documents of the cartel members on agreed price increases). 88 The reliability and accuracy of the analysis depends on whether the analysis has been undertaken rigorously.
Appendix
ESSAY TWO
- ON MARKET POWER AND CARTEL DETECTION: THE FLOUR CARTEL CASE
- Introduction
- Industry background and the flour cartel
- The literature
- Estimating market power and firm conduct in the wheat flour industry
- Detecting cartels
- Conclusion
Appendix
ESSAY THREE
- STRATEGIC ENTRY DETERRENCE: PIONEER FOODS AND THE BREAD CARTEL
- Literature
- Relevant legislation
- The bread industry: A historical background
- Collusion and predation
ESSAY FOUR
- DESIGNING APPROPRIATE REMEDIES AND THE PIONEER FOODS SETTLEMENT
- Designing appropriate remedies in the Pioneer Foods settlement
- A critical assessment of the effectiveness of the Pioneer Foods remedies
- Conclusion
127 Trade unions, NGOs and the National Department for Economic Development supported the Commission. To make a relatively informed assessment of the impact of the discounting agent, I use two comparative approaches. The Commission's investigation revealed that Pioneer Foods and its competitors in the ground wheat and white maize markets (including Tiger Brands, Pioneer Foods, Foodcorp and Premier) were in breach of price-fixing arrangements at various stages from 1998 to at least 2007 of section 4( 1)(b)(i) of the Competition Act.
The R500-million fine represented about 5.6 percent of the turnover of Pioneer Foods' subsidiary Sasko in 2009. The commission was concerned about the effect of the settlement agreement, in the context of the economic recession, that Pioneer Foods could use it to justify job losses and reduced output. I demonstrate the effect of the discount remedy on prices and sales before and after the remedy.
The loss of consumer surplus due to the price surcharge is the sum of the areas A and B. The value of the discount remedy for this person is the sum of the areas S, T and U. The deadweight loss of the discount remedy is shown in Figure 1 as the loss corresponding to area V (see also Appendix 1A for a technical derivation of this result).
With the Pioneer Foods discounter, competing firms did not know the duration and extent of the reductions. The comparative period was defined as the period under review during the implementation of the discount tool. While the total value of the discount agent was known to competitors, the duration and the minimum thresholds were not.
The parameter associated with this dummy variable measures the amount of price reduction (or sales increase) during the discount adjustment. Figures 4-4 and 4-5 show the effect of the discount adjustment on the wholesale price of standard 700 gram loaves of bread. Notes: The table shows estimates of the effect of the discount correction and the standard deviation (in parentheses).
Appendix
ESSAY FIVE
- SETTLING FOR A DISCOUNT: THE PIONEER FOODS PRICE REDUCTION
- Discount remedies and competition law
- The Pioneer Foods discount remedy
- Evaluating the effectiveness of discount remedy in the bread industry
Appendix
CONCLUSION
REFEREED JOURNAL TABLES OF CONTENTS FOR ALL THE PUBLISHED
- ESSAY ONE – Published in the Journal of Industry, Competition & Trade
- ESSAY TWO – Published in the Journal for Studies in Economics and Econometrics. 166
- ESSAY FOUR – Published in the Journal of Competition Law and Economics
- ESSAY FIVE – Published in Agrekon: Agricultural Economics Research, Policy and