The argument of this paper – which can be found here – is straightforward: competition authorities should use a structural remedy when penalising some cartels. The remedy would force cartel member(s) to sell productive assets to other firms for the purpose of making the market more competitive. Given the people the author thanks, and the example he provides, I believe this was inspired by the recent Brazilian experience.
The paper begins with an overview of developments in cartel sanctions over the last 30 years, including: (i) the adoption of leniency programs, (ii) a marked increase in the amount of pecuniary penalties, and (iii) the imposition of criminal sanctions. However, ‘Even if all of these developments have resulted in substantial progress in the fight against cartels, the evidence is that current enforcement falls well short of being an effective deterrent. Many cartels continue to form and operate (…). Furthermore, many of these cartels are not the product of rogue employees but rather are the result of calculated decisions by upper level management. It would seem that collusion remains a viable business strategy in the eyes of many high-level executives’.
The conclusion that the author draws from this: ‘is that a competition authority needs more weapons in its arsenal. (…)The penalty is to require one or more cartel members to divest themselves of assets for the purpose of making the market more competitive (…).Those assets would either be sold to an entrant (e.g., a foreign supplier that is not currently in the market), a non-cartel member (if the cartel was not all-inclusive), or another cartel member. The asset reallocation would be designed to make the market more competitive, by which is meant that collusion is less likely and prices are lower under competition.’
Following this introduction, the paper is structured as follows:
Section 2 describes the benefits of a structural remedy for cartel infringement. The author identifies three main benefits:
- a structural remedy is corrective – by having had a cartel, a market has revealed itself to be predisposed to this type of illegal activity. If the market structure is left intact, collusion could reappear, either as explicit or as tacit collusion. A structural remedy that diminishes the market’s predisposition to collusion is in line with how merger enforcement remedies are applied. It is also argued that divestiture is more compelling for a cartel (where there was collusion) than for a merger (where there may be collusion). Lastly, a structural remedy would be able to deal with instances where (autonomous) parallel behaviour occurs even after a cartel has been dismantled.
- a structural remedy is deterrent – structural remedies can add to financial penalties by making the market more competitive: ‘divestiture that makes the market more competitive means lower profits when firms compete in the post-cartel environment compared to the pre-cartel environment. Like fines and damages, the value of those foregone profits is a financial penalty levied on cartel members’. Furthermore, structural remedies avoid concerns about a firm’s financial viability following the imposition of large pecuniary sanctions.
- A structural remedy is compensatory – A structural remedy ‘benefits consumers because of the lower prices that it delivers. However, it is not focused on benefitting past consumers who purchased during the time of the cartel (as with damages) but rather future consumers who will purchase in the post-cartel period.’ While this does not address the damage suffered by past customers, a structural remedy is compatible with damages actions (which are, in any event, of doubtful use in providing compensation to final consumers in many cases). Furthermore, ‘customer’s demand is persistent over time in many markets for many types of consumers, in which case if they bought during the time of the cartel, they are likely to benefit from the lower post-cartel prices due to divestiture.’
Section 3 describes the costs of this proposal. While a divestiture may offer more benefits than fines and damages, structural remedies are also more costly to implement. First, there are the costs of evaluating the various options by the enforcement bodies, and the resources devoted by the sanctioned companies to discussions about those options. Second, there are transaction costs incurred by the firms when selling and buying the divested assets. Third, there are error costs: ‘economists understand competition more than they understand collusion. For example, shifting capacity from a large firm to a small firm is predicted to lower prices under competition. However, whether that capacity reallocation makes collusion less likely may depend on the entire distribution of capacity within the industry. The error costs associated with determining whether a divestiture will make the market less prone to collusion are probably larger than determining whether prices will be lower under competition .’
As a result, structural remedies should be used only when those error costs are manageable. That depends on the particular market structure and the feasible set of divestitures, which is discussed in section 4.
Section 4 contains a discussion of when and how to use divestitures. A structural remedy should not be pursued unless a reallocation of assets from cartel members to other firms will, with a reasonable level of confidence, make the market more competitive. For this to happen, a number of conditions must be met: (i) there must be assets that can realistically be divested; (ii) the divesture should lead to lower costs, which may be problematic in industries with economies of scale; and (iii) it must be possible to identify purchasers which can lead to lower price-cost margins under competition, and make collusion more difficult. Obvious examples of this are potential new entrants such as foreign suppliers or producers of complementary products.
The author then looks at when it is most appropriate to apply a structural penalty. This is said to be the case when: (i) collusion has lasted for a long time, because this provides evidence that the market is prone to collusion; (ii) when collusion involved senior executives. Senior executives, as opposed to middle managers or sales representatives, will understand the potentially high costs of divestiture. Divestiture is therefore more likely to act as a deterrent in such cases; (iii) when some or all of the cartel members are repeat offenders. Repeat offenders are evidence of a systematic tendency to collude, either at market- or corporation-level, and mean that a restructuring of the corporation or the market is more likely to be warranted; (iv) given jurisdictional limitations, a structural remedy will often have to be limited to domestic cartels.
Section 5 assesses the legality of structural remedies. There are some matters specific to US law that I will skip. However, the main concern is universal: a concern that a structural remedy may involve the unequal treatment of offenders. There may not be an appropriate set of divestitures that would have all cartel members divest a similar percentage of their assets, either because such a collection of divestiture assets does not exist (due to how assets are distributed and structured); or because such a collection of divestiture assets exists but it would inadequately serve the goal of making the market more competitive. According to the author, this does not really pose a problem because it is not that different from joint and several liability or the possibility of applying different fines to different companies. He also notes that the firms which are forced to divest are not necessarily worse off than those that are not, since the effect of the divestiture would be on the market (and hence would affect overall industry profits).
Section 6 provides some examples of cartels for which a structural penalty would seem to be both feasible and appropriate. These include the Gypsum cartel in the US, cement cartels worldwide (and particularly in Brazil, where asset divestment was required in 2014), the retail pharmacies’ cartel in Chile, the hospital divestment order in the UK (which was not based on collusion), and the US investigation into airlines.
This is a very interesting paper (and proposal). My main concern is that it skates over the potential difficulty of enforcing a divestiture remedy. While the costs of such remedies are discussed, they are dealt with rather breezily by arguing that merger control does this all the time. But divestitures in merger control depend largely on firms still wanting to merge despite a divestiture being required, and I’m not sure the mechanism can be easily transposed to the antitrust enforcement sphere. While a company is able to assess whether it is acceptable to enter into such a commitment in the merger control sphere, this assessment will have to be made by the competition authority if divestiture is a sanction. Furthermore, structural remedies are often quite complex – and involve behavioural commitments to ensure their effectiveness, which would impose further costs on the enforcement agency.
Furthermore, it is unclear how such a remedy would interact with leniency applications – would the lenient applicant be exempt from divestiture? –, and which company would be forced to divest its assets – which would raise issues of discrimination and equality of treatment by the competition agency. While this last point is dealt with in section 5, its treatment feels superficial – joint and several liability tends to allow for contribution between the liable parties, and the imposition of disparate penalties needs to be justified by the existence of objective criteria or facts that underpin the differentiated treatment (e.g. leniency applications). None of these mechanisms relate to potential future benefits unconnected with the specific behaviour of the sanctioned company.
Furthermore, the author himself acknowledges that: ‘if a divesting firm does not receive fair market value for its assets – due to insufficient competition among buyers of the assets – then it would be worse off’ – but presents no solution to this. Given that the divestiture would be mandatory, sales below market value are almost guaranteed unless a company is allowed – as it is in merger control – to decide not to sell. Which raises yet another issue: what happens when there are not appropriate buyers or offers in a sufficient amount? And who decides this – the competition agency?
This is not to detract from the paper, or from the value of the proposal. Actually, the author himself acknowledges in the conclusion that: ‘the proposal is modest. It is for divestiture to be included in a competition authority’s array of penalties and that a competition authority actively look for situations to use it.’ This seems right, but it is still worth pointing out that there may be sound reasons why this remedy has not been adopted across the world, and that further work is needed before structural remedies can be widely adopted.