ANTI-COMPETITIVE BEHAVIOUR IN THE AGRI-FOOD AND STEEL VALUE CHAINS IN THE SOUTH AFRICAN MANUFACTURING SECTOR

Despite the fact that it has existed for over ten years, the Competition Act has had little impact in diluting the dominance of big business in the South African manufacturing sector. This study sought to ascertain the extent of anti-competitive behaviour in two sub-sectors of the South African manufacturing sector and to determine whether the competition authorities should focus on supporting SMEs as competitors to big business. The findings indicated that SMEs in these two subsectors face unique difficulties in fighting anti-competitive behaviour, and that there is scope for the competition authorities to facilitate the participation of SMEs in the economy through the use of tools such as market inquiries, the Corporate Leniency Policy and structural remedies.


INTRODUCTION
The legacy of apartheid has resulted in the South African economy being dominated by large businesses, making it difficult for smaller entities to grow (Chabane, Machaka, Molaba, Roberts & Taka, 2003;Hartzenberg, 2006;Kampel, 2005).Since the advent of democracy in 1994, there has been an awareness of the importance of small and medium enterprises (SMEs), and their promotion has become a policy goal of the South African government (Bradford, 2007;Ligthelm, 2008).SMEs are described in the National Small Business Act (Republic of South Africa, 1996a) as those having annual turnovers of up to R50 million and employing up to 200 people.
This study seeks to question the usefulness of competition policy in facilitating the participation of SMEs in two sub-sectors of the South African manufacturing sector -the agrifood sub-sector and the steel sub-sector.Given the financial resources and time needed for SMEs to bring a case against big business (Kampel, 2005), SMEs tend to opt out of markets dominated by big business or simply tolerate anti-competitive behaviour (Chabane et al., 2003;Hartzenberg, 2006) instead of relying on remedies provided for in the Competition Act (Republic of South Africa, 1998), Accordingly, the objectives of the study were to ascertain the extent of anti-competitive behaviour in the agri-food and steel value chains, and to determine whether the competition authorities should protect SMEs as a specific class of competitor.The research questions addressed in this study are the following:  To what extent are SMEs involved in, or victims of, anti-competitive behaviour in the agrifood and steel value chains?
 Given that SMEs lack the resources to use the remedies contained in the Competition Act, should the competition authorities protect SMEs as a specific class of competitor?

Performance of the competition authorities in assisting SMEs
Criticism that the competition authorities were inaccessible increased following the recent rulings of the Competition Appeal Court (CAC) and the Supreme Court of Appeal (SCA).These rulings held that the Competition Commission was not allowed to widen a complaint against one party to include other parties -and so initiate an industry-wide investigation -unless it had the evidence to do so (CAC, 2011).Thus, the SCA ruled that the Competition Commission could not expand a milk cartel investigation against Clover, Parmalat and Nestlé to include two smaller processors, Woodlands Dairy and Milkwood Dairy (SCA, 2010).The SCA ruling was upheld by the CAC (2011), which noted that the Competition Commission could not expand an abuse of dominance case against Sasol to include allegations that other fertiliser producers, Yara SA and Omnia Fertiliser, were colluding with Sasol on fixing fertiliser prices (CAC, 2011).The rulings have been criticised for curtailing the power of the Competition Commission to launch industry-wide investigations (Bleby, 2011;Marais, 2010).
SMEs have also not successfully used the remedies of the Competition Act (Republic of South Africa, 1998) to tackle anti-competitive behaviour, namely horizontal and vertical restricted practices and abuse of dominance (Hartzenberg, 2006).This is because it is difficult to prove collusion in oligopolistic economies where big business can coordinate its behaviour without formal agreements (Albors-Llorens, 2002-2003).The remedies also require proof that the anticompetitive conduct threatens the competitiveness of a market sector, and not only the individual SME (Hartzenberg, 2006;Kampel, 2005).
Despite the aforementioned criticisms, ten years may be too short a period in which to judge the efficacy of the competition authorities.Reflecting on the performance of the Competition Commission in its first decade, Deputy Commissioner Tembinkosi Bonakele explained that the initial focus was to set up the institution to control mergers as opposed to combatting anticompetitive practices (Competition Commission, 2009a (Competition Commission, 2008;2009b).These remedies supplement the authorities' ability to impose structural remedies to dilute the influence of dominant firms in a market (Motta, 2009).

Competition legislation as a means of achieving socio-economic objectives
Reekie (1999) earlier argued that socio-economic objectives can better be achieved through instruments other than competition legislation.In this respect, Reekie (1999) supports a narrow definition of competition policy which is common in the United States.The 1992 United States Department of Justice's Horizontal Merger Guidelines state that the aim of competition policy is economic efficiency and consumer welfare (Chabane et al., 2003;Theron, 2001).This narrow view, known as the Chicago School approach (Cook, 2002;Fox, 2003;Jones & Sufrin, 2011), is tolerant of monopolies if they are justified by economies of scale (Cook, 2002) or if they promote superior performance (Griffith, 2000).The efficient monopoly argument discourages interference in monopolies unless consumer welfare suffers, even if this is at the expense of smaller competitors (Fox, 2007;Jones & Sufrin, 2011).
Contrary to the Chicago School, European Union Competition legislation recognises that a monopoly may be anti-competitive if it restricts SME development (Fox, 2003;Jones & Sufrin, 2011).The European approach, also known as the Ordoliberalist approach, acknowledges that SMEs may need to be protected against monopolies in order to create a more equitable economy (Jones & Sufrin, 2011).It can be argued that this approach is appropriate for developing countries and emerging market economies, which are often dominated by businesses that achieved their dominance through state support (Cook, 2002;Ramburuth & Roberts, 2009).In such economies, unless there is state regulation, big business can retain its power to 'exploit and exclude' (Fox, 2007:229).
However, adopting a pro-SME approach contains threats to competition policy, as it can create the impression that the competition authorities are biased against big business (Jones & Sufrin, 2011).The approach may also endanger consumer welfare if large firms are already competing against each other (Kampel, 2005).Thus, Fox (2003) and Jones and Sufrin (2011) argue that competition policy should protect competition and consumers, but not competitors, even if these competitors are SMEs.

Frameworks to analyse competition in an industry
The lack of public statistics to assess concentration levels in the South African manufacturing sector meant that the present study had to rely on qualitative frameworks to analyse industry competitiveness.Two such frameworks are value chain analysis (Kaplinsky, 2004;Machaka & Roberts, 2003) and Porter's Five Forces Framework (Gamble & Thompson, 2009;Narayanan & Fahey, 2005).
Value chain analysis has its roots in the work of Michael Porter (1979;1980), who linked the profitability of a firm to the firm's environment and its strategy to accommodate this environment.A firm's environment comprises the enterprises with which it interacts, and all the enterprises together generate a value chain of activities needed to create, market, distribute and finally dispose of a product (Kaplinsky, 2004).Porter (1979;1980) and Porter and Kramer (2011) note that there are five ways in which participants in a value chain can impact a firm.These impacts, or forces, determine the level of competition within an industry and the competitive strategies available to a firm.The five forces are explained by Porter's Five Forces Framework (or 'Porter's FFF'), which is set out in FIGURE 1.

FIGURE 1: The five forces driving industry competition
Source: Porter (1980) The usefulness of Porter's FFF in ascertaining the extent of competition in an industry has been criticised for not accommodating the realities of developing countries.Narayanan and Fahey (2005) argue that Porter's FFF assumes a high degree of similarity between rivals in an industry and between rivals and new entrants and ignores the reality that, in developing countries, SMEs operate from a disadvantaged position (Ramburuth & Roberts, 2009).
Value chains are distinguished by their governing procedures.The firm governing the chain controls how other participants enter, innovate or grow in the chain (Altenburg, 2006a;Kaplinsky, 2004;Ponte & Gibbon, 2005).Governance turns a value-chain analysis into an analytical tool (Roberts, 2004), as governance determines how profits are shared in the chain and how the chain evolves (Altenburg, 2006a).Thus, value chain analysis can reveal those parties who inhibit growth in the chain (Barnes & Morris, 2004).
Governance distinguishes buyer-driven chains from producer-driven chains (Ponte & Gibbon, 2005).In buyer-driven chains, buyers (such as food retailers) own and control assets, such as brands and logistical infrastructure that allow them to co-ordinate other players (Barnes & Morris, 2004).In producer-driven chains, upstream participants control key inputs and/or technology (Barnes & Morris, 2004;Altenburg, 2006b).Examples include the steel industry, where the production and pricing of the steel used in manufacturing is controlled by fewer than five players.

The South African manufacturing sector
This study examines two manufacturing sub-sectors classified according to the Standard Industry Classification (SIC) used by the Department of Trade and Industry, namely the food and beverage sub-sector (SIC 300) and the iron and steel and metal products sub-sector (SIC 350).These sub-sectors fall within the Competition Commission's priority sectors, being the Food and Agro-Processing priority sector and the Intermediate and Industrial Products and the Infrastructure and Construction sectors.

The agri-food value chain
The agri-food value chain covers food production from the farm to the consumer.The section of the value chain reflecting the relationship between agro-processors and retailers has been dominated by research that concentrates on the grain value chain.A search of the Business Source Premier and JStor databases revealed that peer-reviewed research published from 1 January 2000 to 30 September 2011 is dominated by Kirsten, Louw and Traub of the University of Pretoria, Jayne of Michigan State University and Roberts and Chabane of the Competition Commission.As the present study did not aim to undertake primary research on value chains, the study concentrated on the existing research, as secondary data, depicted in FIGURE 2.
The highly concentrated nature of food processing and retailing in South Africa is largely attributed to apartheid-era agricultural marketing control boards (Mather, 2005;Ramburuth & Roberts, 2009).These boards restricted entrants into the market (Traub & Jayne, 2008) and so constrained competition (Abu & Kirsten, 2009).The enactment of the Marketing of Agricultural Products Act (Republic of South Africa, 1996b) deregulated the agricultural sector and created expectations that the market would open up for SME processors as had happened in Zimbabwe, Zambia, Mozambique and Kenya (Traub & Jayne, 2008).Instead, an oligopolistic market arose in the maize-milling (Abu & Kirsten, 2009), dairy (Cutts & Kirsten, 2006) and grain-storage (Traub & Jayne, 2008) industries.
While oligopolies per se are not anti-competitive, a market constituted of few firms is conducive to anti-competitive conduct (Jones & Sufrin, 2011).Cutts and Kirsten (2006) confirmed this conclusion for the South African grain and dairy markets, a conclusion later supported by Alemu and Ogundeji (2010).Both sets of researchers found that large food manufacturers and retailers quickly passed on higher commodity prices to consumers, but delayed passing on lower prices (Alemu & Ogundeji, 2010;Cutts & Kirsten, 2006).Abu and Kirsten (2009) conclude that poor profitability restricts the growth of SME millers, as they could not achieve economies of scale.However, in the NAMC studies, the lack of growth of SMEs was attributed to an inability to compete against the large mills that have the capital to ride out grain price fluctuations and engage in price wars with SME millers (Louw et al., 2010a;2010b).Thus, Louw et al. (2010b) conclude that SME participants should be protected against larger millers.

The steel value chain
The  The South African steel value chain is dominated by an upstream sector constituted of steel mills such as Arcelor Mittal SA.This sector is capital-intensive and requires economies of scale.The downstream sector adds value by using steel in manufacturing, mining and construction (Walker & Jourdan, 2003).This sector is fragmented and labour-intensive (Bezuidenhout & Cock, 2009;Roberts & Rustomjee, 2009), and it is also underdeveloped despite producing higher-value products (Walker & Jourdan, 2003).
The commanding role of upstream players in the South African economy can be attributed to the Minerals and Energy Complex (hereafter 'the MEC'), which characterised the South African economy until the 1990s (Bezuidenhout & Cock, 2009).The MEC arose from the apartheid government's requirement for self-sufficiency in the supply of energy and manufacturing inputs (such as steel and chemicals) (Roberts & Rustomjee, 2009).The MEC focus created an industrial policy that favoured capital-intensive mining industries and the manufacturing enterprises based on these industries.The result is that the local economy has not diversified away from mined resources (Fine & Rustomjee, 1996).

Recycling
Scap Metal Dealers

METHODOLOGY
The research design was based on the methodology used by Clarke, Evenett and Lucenti (2005) and entailed constructing a database of newspaper articles on anti-competitive practices in developing countries.The scope of the study was restricted to the aforementioned two subsectors, and the database contained only English-language South African newspaper article and Competition Commission press releases published from 1 January 2005 to 31 March 2011 on the Newsmonitor database.Where media reports lacked sufficient detail, they were cross-checked against the press releases, a practice advocated by Zikmund (2003).This tactic overcame the shortcoming of secondary data not always being accurate (Clarke et al., 2005).
In constructing their database, Clarke et al. (2005) used a predetermined list of anticompetitive practices.Such a procedure was not adopted here, as it yielded too few results on the Newsmonitor database.In addition, all Newsmonitor results referred to instances of anticompetitive behaviour under Competition Commission investigation.Accordingly, the study's database was constructed solely from articles found in a subsidiary Newsmonitor database called 'Competition Commission'.An open coding strategy was adopted, with categories based on the story grammar framework developed by Franzosi (1989;1994;1998;2010a).Franzosi's methodology is suited to analysing newspaper articles (Elliott, 2005).
In analysing each article or press release, every incident of anti-competitive behaviour was categorised into the story grammar structure of who (the subject, or the actor being accused of anti-competitive behaviour), did what (the verb, or the anti-competitive behaviour), to whom (the object complaining of the anti-competitive behaviour), when (the time), where (the subsector of the economy) and why (being the motivation given for the behaviour) (cf.Franzosi, 2010b;Wada, 2005).The codified information was then quantified (cf.Franzosi, 1989;Wada, 2005).

DISCUSSION OF FINDINGS
The objectives of the present study were to ascertain the extent to which SMEs were participants in, or victims of, anti-competitive behaviour in the agri-food and steel value chains and to determine whether the competition authorities should protect SMEs as a specific class of competitor given that they lack the resources to make full use of the remedies contained in the Competition Act.The data is of such a nature that it was deemed appropriate to present the findings and discussion simultaneously.

The agri-food value chain
Eighteen instances of anti-competitive conduct were reported in the agri-food value chain.These cases are listed in APPENDIX A have been summarised on a value-chain diagramme in FIGURE 4. It can be seen that anti-competitive behaviour was not restricted to one area of the value chain, an indicator that the description of the agri-food value chain as buyer-driven (Altenburg ,2006b;Ponte & Gibbon, 2005) is too simplistic.

FIGURE 4: Cases considered in the agri-food value chain
Source: Compiled by authors FIGURE 4 further lists the entities accused of anti-competitive behaviour, which were large manufacturers in 11 instances, and SME manufacturers in four instances.SMEs reported being bullied into participating in cartels in two of the four instances, namely the wheat and maize milling cases.This finding confirms the observation that SMEs need to adopt anti-competitive behaviour if they operate in an industry where such behaviour is common (Iheduru, 2004).While the bullying of SMEs has been identified as common in oligopolistic markets (Dabbah, 2010), SMEs also voluntarily participated in anti-competitive behaviour in the agri-lime, animal feed and broiler cases.These cases provide evidence against an Ordoliberalist approach, as they highlight the danger that protecting competitors such as SMEs can lead to the protection of competitors from competition itself (Fox, 2003).
The prevalence of oligopolies in the agri-food value chain was reported earlier (cf.Abu & Kirsten, 2009;Cutts & Kirsten, 2006).The findings of the present study indicate that oligopolies in the agri-food chain promote anti-competitive behaviour.The predominant type of anticompetitive conduct found in the agri-food chain was exclusionary conduct or behaviour that can only be performed by entities who dominate a market.In addition, it was found that cartels in the agri-food value chain were of the longest duration of cartels noted, a factor attributed to the legacy of state support in that value chain (cf.Mather, 2005;Ramburuth & Roberts, 2009).
The exclusionary conduct cases provide evidence to argue against the adoption of a strict Chicago School approach (Cook, 2002;Fox, 2003;Jones & Sufrin, 2011), which advocates against the competition authorities intervening in an economy to support SMEs.SME manufacturers featured among the complainants who initiated Competition Commission cases in the bread, maize milling and ammonia-based fertiliser cases.The bread cartel case involved an SME bakery that alerted the Competition Commission to the fact that Pioneer Foods was threatening to stop flour supplies to bakeries that did not sell bread at pre-determined prices (Crotty, 2010).In the ammonia-based fertiliser case, two SME fertiliser blenders, Nutri-Flo and Profert, took on Sasol at great cost.In that case, the Competition Tribunal ( 2003) ruled in favour of the SMEs.However, the SMEs lacked the resources to pursue the case when Sasol appealed the ruling.The aforementioned cases illustrate that SMEs cannot compete equally as rivals against big business (Narayanan & Fahery, 2005), especially when they act against entities such as Sasol, which, historically, achieved supremacy through state support (Ramburuth & Roberts, 2009).

The steel value chain
Ten cases of anti-competitive behaviour were reported in the steel value chain during the period 1 January 2005 to 31 March 2011.These cases are noted in APPENDIX B, and have been illustrated in a value-chain diagramme in FIGURE 5. Source: Compiled by authors FIGURE 5 indicates that the majority of cases occurred in industries involved in the production of either primary steel or steel products that are sold into other industries.This is indicative of the producer-driven nature of the value chain, where power is concentrated upstream (Kaplinsky, 2004).
Large manufacturers were accused of anti-competitive conduct in each case considered in the steel value chain.However, SME manufacturers also participated in anti-competitive behaviour in six of the ten cases.Moreover, unlike the cases considered in the agri-food value chain, there was no evidence of SME manufacturers being bullied into collusion.
The participation of SME manufacturers in anti-competitive conduct once again confirms the observation made that SMEs need to resort to such behaviour to survive if they operate in a monopolistic value chain (Iheduru, 2004).An example of this can be seen in the wire products case, where ten SME manufacturers colluded in a buyer's group (known in the hearings as 'the Allens Meshco Group') in order to increase their bargaining power against the vertically integrated steel mills, Cape Gate, Scaw and Arcelor Mittal SA.Using Porter's Five Forces Framework (Porter's FFF), the actions of the Allens Meshco buyers group can be described as set out in FIGURE 6.
Barrier to entry: SME wire manufacturers are unable to purchase economically the raw material needed to make wire

Substitutes
Buyers of steel are numerous, and many are SMEs, hence lack bargaining power Wire is produced from steel, hence no substitutes available Vertically integrated steel mills: produce steel and convert this into wire = bargaining power FIGURE 6: Actions of the Allens Meshco group in counteracting anti-competitive conduct in the wire Industry Source: Compiled by authors FIGURE 6 indicates how a barrier to entry to wire manufacturers was created by steel mills selling wire from steel they had produced.The steel buyers were too numerous to bargain on prices and so created a buyers' group.This action did not remedy the anti-competitive structure of the industry, as barriers to entry remained and SME buyers that were not part of the group did not benefit from the new buying power.
Save for one case, namely the wire-rod case, SME manufacturers were not among the complainants who triggered the Competition Commission inquiries in the steel value chain.However, as is the case in the agri-food value chain, SME manufacturers benefited from the cases initiated by big business or by the Competition Commission.Noteworthy is the excessive pricing steel case, as steel is a key manufacturing input for downstream SMEs.While the Competition Tribunal ruled against Arcelor Mittal SA in this instance, it was remitted back to the Competition Tribunal in 2007 by the appeal court with an order that the Tribunal reconsider its excessive pricing determination.The matter remains unresolved, and the case has been used to criticise the efficacy of competition policy in curbing the abuse of dominance in the steel industry (Bezuidenhout & Cock, 2009).

The need for competition authorities to focus on SMEs
At the heart of this study is the question relating to competition policy, namely whether it is sufficient for competition policy to concern itself with protecting competition as a process, or whether there is also a need for such policy to protect individual categories of competitors such as SMEs (Fox, 2003;Jones & Sufrin, 2011).The cases considered have shown that both big business and SMEs were victims of anticompetitive conduct.However, the experiences of SMEs in tackling this conduct are unique.SMEs generally lack the resources to tackle anti-competitive behaviour or to sustain an action once it has been initiated.SMEs are also bullied into participating in anti-competitive behaviour.
The focus on priority sectors has enabled the competition authorities to initiate investigations into anti-competitive behaviour in industries where SMEs are likely to be affected by such behaviour.The authorities have also used the Corporate Leniency Policy to investigate and prosecute cartels that are inimical to SME development.These investigations have saved SMEs the time and cost of fighting such cartels.The competition authorities have also embarked on structural remedies to dilute the power of dominant companies.Thus, in the ammonia-fertiliser case, a structural remedy was imposed on Sasol that, among other things, compelled Sasol to sell its fertiliser blending units to entities approved of by the Competition Commission.The aim of this remedy was to dismantle vertical integration by Sasol in the fertiliser industry and to create opportunities for SMEs to enter the industry.Using Porter's Five Forces Framework, the impact of the remedy on the industry is demonstrated in FIGURE 7. The structural remedy allowed the competition authorities to intervene in the South African fertiliser industry.This interventionist role is in keeping with the Ordoliberalist approach, which requires the competition authorities to consider not only consumer welfare, but also the barriers preventing new entrants into the market (Kampel, 2005).

Barrier to entry
In summary, the study highlighted that by using tools such as market inquiries, the Corporate Leniency Policy and structural remedies, the authorities can facilitate the participation of SMEs in the economy without making this a specific focus of intervention.However, as noted earlier, it may be premature to judge the performance of the competition authorities in assisting SMEs (Competition Commission, 2009a).Accordingly, it may also be premature to recommend whether the South African competition authorities should adopt a more interventionist approach in achieving the objectives set out in the Competition Act (Republic of South Africa, 1998, Section 2) of facilitating the participation of SMEs in the South African economy.

Limitations of the study and areas for future research
The study is limited to the manufacturing sector of the South African economy, and, in particular, to only the agri-food and steel value chains.Accordingly, the study does not provide a complete picture of all anti-competitive activity in the South African manufacturing sector, and a more comprehensive study would add value to this discourse.
The database used in this study contained only articles from the Newsmonitor database entitled 'Competition Commission'.Thus, only cases under Commission inquiry, as reported by the press, were investigated.Accordingly, instances of anti-competitive behaviour which had not yet come to the attention of the authorities were not included.Therefore, there could be more anticompetitive practices in the agri-food and steel value chains than are reported on in this study.
Little academic research has been conducted on the extent of anti-competitive behaviour in the agri-food and steel value chains, and that which does exist has been produced by a small number of researchers.In the agri-food value chain, in particular, peer-reviewed research has focused only on the grain value chain.Deepening academic study relating to these value chains will generate a more accurate view of how entrenched anti-competitive behaviour has become in these value chains.
In ascertaining whether the competition authorities should focus specifically on assisting SMEs, it is necessary to consider the tools which the competition authorities can use to create a more competitive environment without having a specific focus on SMEs.These tools, which include market inquiries, the Corporate Leniency Policy and structural remedies, are fairly new, and research on their impact in facilitating the participation of SMEs in the economy has yet to build up.Accordingly, it is recommended that the impact of the use of these tools for SME development be monitored and further researched.

CONCLUSION
The study sought to answer the following questions:


To what extent are SMEs involved in, or victims of, anti-competitive behaviour in the agri-food and steel value chains?


Given that SMEs lack the resources to use the remedies contained in the Competition Act, should the competition authorities protect SMEs as a specific class of competitor?
The study has indicated that SMEs were both victims of and voluntary participants in anticompetitive behaviour in both the agri-food and steel value chains in the South African manufacturing sector.The participation of SMEs in such behaviour suggests that the competition authorities should not focus specifically on SMEs, because there is the danger that, in doing so, they could protect SMEs from competition itself.However, it was also found that SMEs face unique difficulties in bringing cases to the attention of the competition authorities.SMEs have been spared the cost and time of fighting cases of anti-competitive behaviour by the competition authorities initiating investigations in industries which are likely to be affected by anti-competitive behaviour, and the authorities have used tools such as the Corporate Leniency Policy and structural remedies to tackle cartels, which are inimical to SME development.These tools are still new, and further research into their use by the competition authorities is recommended in order to determine whether the competition authorities should adopt a more interventionist approach to facilitate the participation of SMEs in the economy.

FIGURE 2 :
FIGURE 2: Diagrammatic representation of the South African agri-food value chain Source: Adapted from Louw, Geyser, Trotskie, Van der Merwe, Scheltema and Nicholson (2010a) A search from 1 January 2000 to 31 August 2011 of the Business Source Premier and JStor databases revealed that Abu and Kirsten (2009) produced the only published academic study on the competitiveness of small South African maize millers.In addition to this peer-reviewed study, there were also two non peer-reviewed studies commissioned by the National Agricultural Marketing Council (hereafter 'the NAMC'), namely those undertaken by Louw, Geyser, Trostskie, Van der Merwe, Scheltema and Nicholson (2010a) and Louw, Geyser and Schoeman (2010b).

FIGURE 3 :
FIGURE 3: Diagrammatic representation of the South African steel value chain Source: The World Steel Association (2008) A search of the Business Source Premier and Jstor databases for peer-reviewed articles published from 1 January 2000 to 31 August 2011 identified little academic research on anticompetitive behaviour in the South African steel value chain.The research that does exist is dominated by Roberts and Chabane of the Competition Commission.

Figure 5 :
Figure 5: Cases considered in the steel value chain

FIGURE 7 :
FIGURE 7: Effect of the structural remedy in the Sasol ammonia fertiliser case Source: Compiled by authors shattered: the sale of of 5 Sasol blending plants allows for at least 5 new entrants unrelated to dominant players ).However, in 2006, the Competition Commission took the strategic decision to proactively tackle anti-competitive behaviour(Competition Commission, 2009a), and identified priority sectors for investigation, including Food and Agro-Processing, Intermediate Industrial Products and Infrastructure and Construction(Competition Commission, 2010).The Competition Amendment Act (Republic of South Africa, 2009) also introduced new remedies: for example, the market inquiry provision which allows the Competition Commission to initiate investigations into anti-competitive behaviour, and the Corporate Leniency Policy, which grants immunity from prosecution to cartel members who disclose cartel activities Competition Commission (2009)prioritised the Intermediate Industrial Products and the Construction and Infrastructure sectors for investigation.The steel value chain falls into both sectors, as steel is a manufacturing input and a construction building material.A representation of the steel value chain is set out in FIGURE3.