For more information on how to contribute to the series, contact the working paper editor, Wihan Marais.
Working papers – 2025
Competition Policy in South Africa: From 1994 to Now. Two competing approaches in South African competition policy: merger control and anti-cartel enforcement over three decades (WPS01/2025)
Prof. Willem H. Boshoff
This paper examines the evolution of South African competition policy since the Competition Act of 1998, focusing on merger control and anti-cartel enforcement. Using a Bayesian decision-theoretic framework, the analysis evaluates whether enforcement trends are driven by research and policy experience or by shifting policy preferences. The findings reveal two divergent paths: merger control has become significantly more interventionist with broader theories of harm, largely driven by changing policy preferences rather than empirical evidence; meanwhile, anti-cartel enforcement expanded rapidly before stabilising, with its growth primarily supported by learning effects and international practices, though recent novel interpretations of collusion also reflect preference shifts. The paper identifies tensions between these approaches—one evolving based on economic literature and international precedent, the other shaped by policymaker preferences. It concludes that competition policy dominated by preferences rather than systematic, evidence-based evolution may undermine effective competition enforcement, even when pursuing broader objectives beyond economic efficiency.
Capacity Collusion and Overcharge Estimation (WPS02/2025)
Prof. Willem H. Boshoff and Dr. Thomas Fagart
Collusion may change price dynamics prior to and, especially, following a formal period of coordination. This paper presents a dynamic model of capacity choice under partial irreversibility to study the implications of a capacity cartel on price dynamics and damage estimates. We study both transitory pre- and post-cartel adjustment effects as well as a long-run effect and find that collusion damage continues after the cartel ends, and may persist indefinitely when demand remains constant. These effects imply biased estimates of cartel overcharge by both dummy-variable and forecasting methods, both of which rely on data during and after the collusion.
Sustainability, Competition, and Decision-Making Under Uncertainty: A Bayesian Approach to Green Merger Analysis in South Africa (WPS03/2025)
Prof. Willem H. Boshoff and Katherine Stainton
This paper develops a Bayesian decision-making framework to assist South African competition authorities in evaluating anti-competitive mergers that generate environmental efficiencies. As the urgency of climate change and environmental sustainability grows, competition regulators increasingly confront the complex task of evaluating green efficiencies claimed in merger transactions while maintaining their core mandate of protecting competition. South Africa’s competition regime, which explicitly incorporates public interest considerations alongside traditional competitive assessments, provides a unique context for examining this issue. Building on Boshoff’s (2014) application of Bayesian techniques to market definition, this research extends the probabilistic treatment of classification errors to merger review decisions. Through theoretical development and empirical application to the Pioneer/Pannar Seeds merger case, this research demonstrates how Bayesian decision theory can provide a more structured and transparent approach to evaluating environmental efficiencies under uncertainty. The framework offers practical tools for competition authorities navigating the challenges of integrating sustainability considerations into merger analysis while contributing to broader academic discussions about the role of competition law in advancing environmental objectives.
Working papers – 2022
Samantha Kee and Yongama Njisane
There has been a growing interest by competition authorities and antitrust scholars on common ownership linkages by institutional investors which may be leading to a softening of competition in markets. This paper offers a glimpse into the potential extent and nature of common ownership in South Africa based on a select number of large institutional investors. From our limited observations on common ownership, we find that common ownership tends to occur with rival incumbent firms in markets. We also observe that our selected large institutional investors, as a group (including the holding firm and subsidiaries), tend to hold minority shares in parent firms with several diversified product portfolios of their own. The main purpose of this paper is to stimulate further research on whether common ownership by large investors may potentially result in softer competition in industries. As a natural starting point, research should also consider further determination of the extent to which common ownership occurs in South African industries. Economic literature would also benefit from an exploration of whether there is any relationship between firm behaviour and common ownership and the nature of such a relationship, especially in the context of South Africa.
Working papers – 2021
Prof. W.H. Boshoff, Prof. L.M. Froeb, Roan Minnie, Prof. S. Tschantz
Vertical merger models are complex systems built on (i) a network of, e.g., upstream sellers and downstream retailers (ii) who bargain bilaterally in the presence of externalities (iii) created by competition between downstream retailers (iv) facing a consumer demand surface. We simulate the effects of vertical mergers in seven different models to identify which, and by how much, various assumptions about the nature of bargaining lead to anticompetitive outcomes. We find, among other things, that:
- Compared to derived demand models, bargaining models increase the scope for anticompetitive outcomes because they reduce upstream margins which reduces the benefit of merger, the elimination of double marginalization.
- Nash-in-Nash and Nash-in-Shapley can give opposing predictions about whether vertical mergers are anticompetitive.
- In a 2×1 network, vertical merger does not affect output, though they do redistribute profit by giving the integrated firm a better threat point.
- The differences between Nash-in-Shapley and Nash-in-Nash disappear, as there is only one instrument to both split the pie and increase its size.
- A smaller aggregate elasticity (less substitution between inside and outside goods), and a stronger nest (more substitution among inside goods) increases the likelihood of an anticompetitive vertical merger.
Because the assumptions about the nature of bargaining can predetermine outcomes, it is critical to ensure that a model captures the significant features of pre-merger competition, and the loss of such competition following merger. However, this may be difficult to do because many of the assumptions about bargaining–critically, the alternatives to agreement-are unobservable.
Prince Changole & Willem Boshoff
Merger control regimes in various jurisdictions, especially in Africa, feature non-competition objectives in addition to conventional goals, such as the maintenance or promotion of competition. Such `public interest’ objectives, including the promotion of employment, small business and particular industries, create special challenges for competition authorities. Furthermore, the broad definition of, and complexity associated with, non-competition objectives may increase uncertainty about merger control. We study the systematic impact of public interest concerns on South African merger decisions, in terms of duration of adjudication and consistency over time. Our results suggest that the adjudication of mergers featuring public interest concerns take longer. More importantly, these cases have a higher probability of having conditions imposed for approval and this has been increasing steadily over the past decade. This indicates more aggressive merger control and raises policy questions about the consistency, and hence predictability, of South African merger decisions.
Working papers – 2020
Prof. W.H. Boshoff
The declaration of a state of national disaster in South Africa, due to the Covid-19 pandemic, was followed by excessive-pricing regulations pertaining to certain consumer and medical products and services. The regulations and their application suggest an intertemporal benchmark to judge excessive pricing, deviating from previous practice. Intertemporal comparisons assume a structural shift during Covid-19 that changes competitive conditions, related to changes in consumer behaviour. Such comparisons must also account for demand and cost changes. While the Covid-19 regulations allow for cost-based price increases, demand-based increases are not explicitly accounted for, suggesting that the regulations are framed more generally as price-gouging regulations. The differences between price-gouging and excessive pricing benchmarks depends on the type of disaster-period demand shock. They are similar following a transitory demand spike, provided sufficient time is allowed for dynamic price behaviour, but differ markedly when demand is elevated for the duration of the disaster period. Applying simple cost-based comparisons in recently concluded cases against smaller retailers are consistent with excessive pricing, given the presence of a demand spike. To the extent that these involve persistently-higher demand, cases against wholesalers and larger retailers will be more complicated, as such demand must be reflected in competitive prices.
Working papers – 2019
Paul W.J. de Bijl and Helanya Fourie
This paper explores the relationship between local public ownership and the sustainability mission of companies in the energy transition. Public ownership may matter due to “contractual incompleteness” where legislation and regulation fall short in dealing with market failures or addressing public interests. Ownership provides a measure of influence and control over the mission and strategy of a company, which is important in light of the unanticipated consequences of the energy transition. The role of municipal ownership is underlined by the growing popularity of district heating and cooling systems and the impact of the energy transition on the urban environment.
Working papers – 2017
Under imperfect competition, firm’s incentive to deviate from a collusive agreement is usually short term: the rm increases its immediate prot but reduces its future prots due to the upcoming competitors’ reaction. When firms have to invest to increase their production, and that investment is irreversible, a rm may deviate to preempt its competitors and obtain a dominant position on the market. When firms are more patient, preemption is more protable, and collusion may thus be harder to sustain. This result, contrasting with the literature on collusion, suggests that folk theorems may be inappropriate to study collusion in Markovian frameworks.
Willem Boshoff & Johannes Paha
Firms sometimes violate competition laws by agreeing on increases of list prices. The economic effects of such list price collusion are far from clear because the cartel firms might deviate secretly from the elevated prices by granting their customers discounts. This article presents case evidence suggesting that agreements on list prices are not infrequently observed in cartel cases. It also reviews theoretical, empirical, and experimental literature in economics showing under what conditions such list price collusion causes the discounted transaction prices to rise. This is relevant for competition authorities in developing a theory of harm when prosecuting cartels, and also for the customers of the cartel firms when suing the conspirators for the repayment of damages.
Recurrent collusion: Cartel episodes and overcharge in the South African cement market (WPS03/2017))
Willem Boshoff & Rossouw van Jaarsveld
Collusion is often a recurrent phenomenon, with cartel periods interspersed by periods of greater competition. In developing countries in particular, industries with a history of legal collusion are often characterized by recurrent collusion. Canonical models implicitly treat collusion as recurrent by modelling collusion as a state dependent outcome, often based on an unobserved demand state. Yet empirical studies have paid less attention to recurrent collusion. This paper proposes a Markov regime-switching (RS) model to detect recurrent periods of collusive damages and to estimate price overcharge in these cases. An empirical model of recurrent collusion must satisfy three properties. Firstly, it should account for dierent datagenerating processes during collusive and non-collusive episodes. Secondly, it should be able to detect the dates and duration of collusive and non-collusive episodes. Thirdly, it should account for exible transitions between collusive and non-collusive episodes. We argue that the RS model meets these requirements. We demonstrate these features in an application to the South African cement market, which, similar to cement markets in a number of other countries, have experienced recurrent collusion. Our results suggest that the exclusion of non-collusive periods { including price wars { yield higher overcharge estimates than conventional approaches. In competition policy, the RS model can act as a screening tool, to identify recurrent collusive behaviour. Courts may also nd the RS model useful when estimating collusive damages, especially in private litigation, where the court must also determine the period of liability.
Working papers – 2022
Samantha Kee and Yongama Njisane
There has been a growing interest by competition authorities and antitrust scholars on common ownership linkages by institutional investors which may be leading to a softening of competition in markets. This paper offers a glimpse into the potential extent and nature of common ownership in South Africa based on a select number of large institutional investors. From our limited observations on common ownership, we find that common ownership tends to occur with rival incumbent firms in markets. We also observe that our selected large institutional investors, as a group (including the holding firm and subsidiaries), tend to hold minority shares in parent firms with several diversified product portfolios of their own. The main purpose of this paper is to stimulate further research on whether common ownership by large investors may potentially result in softer competition in industries. As a natural starting point, research should also consider further determination of the extent to which common ownership occurs in South African industries. Economic literature would also benefit from an exploration of whether there is any relationship between firm behaviour and common ownership and the nature of such a relationship, especially in the context of South Africa.
Working papers – 2022
Samantha Kee and Yongama Njisane
There has been a growing interest by competition authorities and antitrust scholars on common ownership linkages by institutional investors which may be leading to a softening of competition in markets. This paper offers a glimpse into the potential extent and nature of common ownership in South Africa based on a select number of large institutional investors. From our limited observations on common ownership, we find that common ownership tends to occur with rival incumbent firms in markets. We also observe that our selected large institutional investors, as a group (including the holding firm and subsidiaries), tend to hold minority shares in parent firms with several diversified product portfolios of their own. The main purpose of this paper is to stimulate further research on whether common ownership by large investors may potentially result in softer competition in industries. As a natural starting point, research should also consider further determination of the extent to which common ownership occurs in South African industries. Economic literature would also benefit from an exploration of whether there is any relationship between firm behaviour and common ownership and the nature of such a relationship, especially in the context of South Africa.
Working papers – 2021
Prof. W.H. Boshoff, Prof. L.M. Froeb, Roan Minnie, Prof. S. Tschantz
Vertical merger models are complex systems built on (i) a network of, e.g., upstream sellers and downstream retailers (ii) who bargain bilaterally in the presence of externalities (iii) created by competition between downstream retailers (iv) facing a consumer demand surface. We simulate the effects of vertical mergers in seven different models to identify which, and by how much, various assumptions about the nature of bargaining lead to anticompetitive outcomes. We find, among other things, that:
- Compared to derived demand models, bargaining models increase the scope for anticompetitive outcomes because they reduce upstream margins which reduces the benefit of merger, the elimination of double marginalization.
- Nash-in-Nash and Nash-in-Shapley can give opposing predictions about whether vertical mergers are anticompetitive.
- In a 2×1 network, vertical merger does not affect output, though they do redistribute profit by giving the integrated firm a better threat point.
- The differences between Nash-in-Shapley and Nash-in-Nash disappear, as there is only one instrument to both split the pie and increase its size.
- A smaller aggregate elasticity (less substitution between inside and outside goods), and a stronger nest (more substitution among inside goods) increases the likelihood of an anticompetitive vertical merger.
Because the assumptions about the nature of bargaining can predetermine outcomes, it is critical to ensure that a model captures the significant features of pre-merger competition, and the loss of such competition following merger. However, this may be difficult to do because many of the assumptions about bargaining–critically, the alternatives to agreement-are unobservable.
Prince Changole & Willem Boshoff
Merger control regimes in various jurisdictions, especially in Africa, feature non-competition objectives in addition to conventional goals, such as the maintenance or promotion of competition. Such `public interest’ objectives, including the promotion of employment, small business and particular industries, create special challenges for competition authorities. Furthermore, the broad definition of, and complexity associated with, non-competition objectives may increase uncertainty about merger control. We study the systematic impact of public interest concerns on South African merger decisions, in terms of duration of adjudication and consistency over time. Our results suggest that the adjudication of mergers featuring public interest concerns take longer. More importantly, these cases have a higher probability of having conditions imposed for approval and this has been increasing steadily over the past decade. This indicates more aggressive merger control and raises policy questions about the consistency, and hence predictability, of South African merger decisions.
Working papers – 2020
Prof. W.H. Boshoff
The declaration of a state of national disaster in South Africa, due to the Covid-19 pandemic, was followed by excessive-pricing regulations pertaining to certain consumer and medical products and services. The regulations and their application suggest an intertemporal benchmark to judge excessive pricing, deviating from previous practice. Intertemporal comparisons assume a structural shift during Covid-19 that changes competitive conditions, related to changes in consumer behaviour. Such comparisons must also account for demand and cost changes. While the Covid-19 regulations allow for cost-based price increases, demand-based increases are not explicitly accounted for, suggesting that the regulations are framed more generally as price-gouging regulations. The differences between price-gouging and excessive pricing benchmarks depends on the type of disaster-period demand shock. They are similar following a transitory demand spike, provided sufficient time is allowed for dynamic price behaviour, but differ markedly when demand is elevated for the duration of the disaster period. Applying simple cost-based comparisons in recently concluded cases against smaller retailers are consistent with excessive pricing, given the presence of a demand spike. To the extent that these involve persistently-higher demand, cases against wholesalers and larger retailers will be more complicated, as such demand must be reflected in competitive prices.
Working papers – 2019
Paul W.J. de Bijl and Helanya Fourie
This paper explores the relationship between local public ownership and the sustainability mission of companies in the energy transition. Public ownership may matter due to “contractual incompleteness” where legislation and regulation fall short in dealing with market failures or addressing public interests. Ownership provides a measure of influence and control over the mission and strategy of a company, which is important in light of the unanticipated consequences of the energy transition. The role of municipal ownership is underlined by the growing popularity of district heating and cooling systems and the impact of the energy transition on the urban environment.
Working papers – 2017
Under imperfect competition, firm’s incentive to deviate from a collusive agreement is usually short term: the rm increases its immediate prot but reduces its future prots due to the upcoming competitors’ reaction. When firms have to invest to increase their production, and that investment is irreversible, a rm may deviate to preempt its competitors and obtain a dominant position on the market. When firms are more patient, preemption is more protable, and collusion may thus be harder to sustain. This result, contrasting with the literature on collusion, suggests that folk theorems may be inappropriate to study collusion in Markovian frameworks.
Willem Boshoff & Johannes Paha
Firms sometimes violate competition laws by agreeing on increases of list prices. The economic effects of such list price collusion are far from clear because the cartel firms might deviate secretly from the elevated prices by granting their customers discounts. This article presents case evidence suggesting that agreements on list prices are not infrequently observed in cartel cases. It also reviews theoretical, empirical, and experimental literature in economics showing under what conditions such list price collusion causes the discounted transaction prices to rise. This is relevant for competition authorities in developing a theory of harm when prosecuting cartels, and also for the customers of the cartel firms when suing the conspirators for the repayment of damages.
Recurrent collusion: Cartel episodes and overcharge in the South African cement market (WPS03/2017))
Willem Boshoff & Rossouw van Jaarsveld
Collusion is often a recurrent phenomenon, with cartel periods interspersed by periods of greater competition. In developing countries in particular, industries with a history of legal collusion are often characterized by recurrent collusion. Canonical models implicitly treat collusion as recurrent by modelling collusion as a state dependent outcome, often based on an unobserved demand state. Yet empirical studies have paid less attention to recurrent collusion. This paper proposes a Markov regime-switching (RS) model to detect recurrent periods of collusive damages and to estimate price overcharge in these cases. An empirical model of recurrent collusion must satisfy three properties. Firstly, it should account for dierent datagenerating processes during collusive and non-collusive episodes. Secondly, it should be able to detect the dates and duration of collusive and non-collusive episodes. Thirdly, it should account for exible transitions between collusive and non-collusive episodes. We argue that the RS model meets these requirements. We demonstrate these features in an application to the South African cement market, which, similar to cement markets in a number of other countries, have experienced recurrent collusion. Our results suggest that the exclusion of non-collusive periods { including price wars { yield higher overcharge estimates than conventional approaches. In competition policy, the RS model can act as a screening tool, to identify recurrent collusive behaviour. Courts may also nd the RS model useful when estimating collusive damages, especially in private litigation, where the court must also determine the period of liability.