OmniCLES is registered with the Dutch Chamber of Commerce under no. 59977078. Copyright 2014 OmniCLES. All rights reserved.
Type hier je alinea.
Type hier je alinea.
Below is a selection of short articles posted on my Linkedin profile. The most recent contributions are posted higher.
The purpose and equality of merger control procedures
Dr. Rob van der Laan, 19 November 2018
This article is a summary of my presentation given at the 2018 ICN Merger Workshop in Tokyo on 8 November. Please note that the presentation and this summary should be taken as a starting point for discussion: I merely observe a potential issue, but I do not profess to have the best answer to resolve this issue. Given its purpose, this summary by necessity brief and does not give an exhaustive review of the issue.
1. Modern competition law is firmly based on economic theory. The market mechanism provides the solution for the great majority of situations. Only some very extreme situations require intervention of the competition authority, for example where competitors engage in customer allocation or when a monopolist abuses its market power.
2. Within competition law, the enforcement of merger control is quite distinct from the enforcement of antitrust legislation. In antitrust, enforcement is ex post following an infringement, be it an anti-competitive agreement or an abuse of a dominant position in trade. In merger control, enforcement is ex ante, predicting the creation or enforcement of a dominant position in trade or a substantial lessening of competition.
3. The nature of merger control - intervening in market behaviour that could reduce competition in the near but still uncertain future - would logically imply that intervention in relation to proposed mergers would be rare. This is reflected by the statistics of the competition authorities.
4. Even if a proposed merger would result in a substantial lessening of competition (or similar threshold required for intervention), this does not automatically imply that the merger is blocked or requires remedies. This is because of the so-called efficiency defence. If the merger results in specific benefits to society that outweigh the negative effects on competition and that cannot be achieved without the merger, the merger will be allowed to proceed.
5. The central proposition of this contribution is that merger control can become biased because non-competition efficiencies and non-competition “inefficiencies” are treated differently in the merger control process. This may result in a sub-optimal solution, both from a total welfare perspective and from a fairness perspective.
6. The merger control process by a competition authority in the jurisdictions that I am most familiar with (the EU, the Netherlands and the United Kingdom) allows for an efficiency defence. The merging parties can propose to the competition authority that, despite a reduction in competition (e.g. the creation or enhancement of a dominant position), the merger should be allowed to proceed because of benefits to society that result of the merger. On the other hand, once a competition authority concludes that there is no substantial lessening of competition, third parties objecting to the merger do not have a similar opportunity to ask for intervention for non-competition reasons.
7. For example, if a merger would result in an SLC but also substantial benefits for the natural environment these benefits can potentially qualify as an efficiency defence. On the other hand, if a merger would not result in an SLC but would also result in significant environmental damage, there is no opportunity to raise such inefficiencies as part of the merger proceedings. This creates a bias in merger control.
8. In addition to non-economic effects of a merger such as quality, there is also a very specific cost (or, more particular, risk) associated with merger control. That is the possibility that a merger may fail. This is a risk inherent in every merger. From a competition perspective, the counterfactual situation without a merger will generally be superior to a post-merger situation that results in exit from the market.
9. As many markets have a great number of suppliers and surmountable barriers to entry, the reduction of competition from a failed merger is not a major issue and inherent of the “right to fail” in a market economy. However, if the service provided by the merger parties is healthcare one may want to re-evaluate the conventional wisdom that consideration of the ex-ante chances of success of a merger should be of no concern to the competition authority. In other words: If a toy shop closes following a failed merger, the costs of society may be very limited. On the other hand, if a hospital closes following a failed merger, the impact on society may be considerable. There appears to be little scope for competition authorities to take such concerns into account once the authority has concluded that there will be no significant lessening of competition.
10. I will highlight this bias by comparing two recent hospital mergers, one in the Netherlands and one in the United Kingdom. Both involve the leading hospitals in the area – respectively the (only) two academic hospitals in the Amsterdam region and the two leading hospitals in the Manchester region.
11. A big difference in the material assessment of the proposed mergers is the market definition: In the UK, each speciality is a relevant product market. The parties respectively offered 84 and 56 specialities in relation to elective and maternity services and 89 and 31 specialised services. The UK competition authority CMA concluded that the merger would result in an SLC in relation to several of these specialities. In the Netherlands, all the health care services are grouped in two large categories of “basic care” and “top care”. Given these much larger relevant product markets, there are many more suppliers and the Dutch competition authority ACM concluded that the merger would not result in an SLC.
12. One may have preferences which of the market definitions adopted is better – a precautionary approach would imply making the assessment at the narrowest relevant product market, and if there would be no competition concerns on the basis of the narrow market definition one can generally conclude that there should be no competition concerns if the relevant product market should be defined more broadly -, the main point I want to make in this contribution is how efficiencies and inefficiencies were taken (or not taken) into account.
13. The conclusion from the UK CMA was that the merger would result in a substantial lessening of competition. This conclusion procedurally opened the way for an efficiency defence. The UK phase II decision noted that “we have been struck in our inquiry by the widespread support of the merger of those NHS-related bodies who we have spoken to including the GMHSCP, CCGs, NHS England, and other providers. We have also been struck by the enthusiasm and support for the merger of each party’s clinical staff we have met. They have all cited the benefits of the merger to patients as their reason for supporting the merger. (..) we have given material weight to the reduction in mortality, and complications and morbidity for a significant number of patients which are likely to result from the merger, which we consider to be extremely significant benefits, in addition to the merger’s likely beneficial impact on patient access and on the hospital experiences for a significant number of patients (..) We consider that the adverse effect likely to result from the SLC that we have found in NHS elective and maternity services and NHS specialised services is substantially lower than the beneficial impact of the RCBs that would be lost as a result of a prohibition remedy (..) Accordingly, we have decided that it would be disproportionate to prohibit the merger. Therefore, we are clearing the merger.” In conclusion, following the SLC established by the CMA the authority was still able to clear the merger as a result of efficiencies.
14. The conclusion from the Dutch ACM was that there was no substantial lessening of competition. Therefore, there was no opportunity to review non-competition efficiencies and risks. The Dutch phase II decision noted that (my translation) “both a number of third parties (RL – I assume the purchasers, i.e. the health care insurers) and the Dutch Healthcare Authority indicate that the proposed merger will result in a very large organisation with the associated risks (for example, whether this organisation is still manageable. This risks however are not within the scope of the merger control assessment and therefore cannot be addressed by the ACM. (..) The Dutch Healthcare Authority predicted risks both in relation to prices and in relation to quality. (..) The Dutch Healthcare Inspectorate stated that it did not have information or means to determine with certainty that the merger would have negative effects on quality and safety of care. However, is noted that a merger process between healthcare provides always results in risks in relation quality and safety of care. (..) The market power of the parties in relation health care in which they respectively already have market power could increase, which can result in higher prices or lower quality in healthcare for which they are in competition. The merger will result in such a large organisation that the Dutch Healthcare Authority foresees a very significant challenge for the parties to manage this organisation, both during the merger process and afterwards. The Dutch Healthcare Authority assessed that there is a risk that this will have a negative effect on the quality of care. The Dutch Healthcare Authority also assessed that the merger could result in an organisation that is too big to fail. The Dutch Healthcare Authority concludes that it is not convinced that public interests are served with the merger.” In conclusion, as the ACM concluded that there was the proposed merge would not result in an SLC, the authority was not able to deal with the risks identified by third parties.
15. I do not argue for a total-welfare analysis of mergers in general. Such would run counter to the aforementioned fundamentals of merger control. In addition, the costs of such a system of merger control is almost certain to outweigh the benefits given the very large percentage of mergers that pose not issue whatsoever from a competition law perspective. However, where the competition authority is empowered to review an efficiency defence, it may - at least for certain sectors and at least for certain risks– be justified to allow the competition authority to assess such risks irrespective of whether the proposed merger would result in a substantial lessening of competition or not.
(The handout can be found at https://icnmergerworkshop2018.jftc.go.jp/agenda.html)
Abuse of dominance and market definition - A comment on the CMA Phenytoin Sodium excessive pricing case CE/9742-13
Dr. Rob van der Laan, 17 July 2017
On 15 June 2017, the CMA published the full text of its infringement decision, finding that Pfizer and Flynn Pharma abused their respective dominant positions by pricing excessively. Despite (or perhaps because?) its 465 pages, the decision makes for interesting reading. This is not only for the relative rarity of excessive pricing cases. In this short comment, I will focus on market definition and the evidencing of the relevant product market.
Some background: The product involved is phenytoin sodium (“PS”), a prescription drug to treat epilepsy. PS is an 80-year old medicine and no longer recommended for new patients. It is estimated that about 10% or 48,000 UK patients are reliant on PS capsules.
From September 2012, Pfizer produced PS capsules were distributed to pharmacies through Flynn rather than directly by Pfizer. The prices for 100mg capsules increased in September 2012 by over 2,200% (of which over 1,300% from the price charged by Pfizer to Flynn). Competitor NRIM started manufacture of PS capsules in April 2013.Between April 2013 and November 2013, NRIM’s SP capsules were significantly cheaper to dispense than Pfizer-Flynn’s product.
The Medicines and Healthcare Products Regulatory Agency (‘MHRA’) recommended in November 2013 that patients who are stabilised on a particular manufacturer’s phenytoin sodium capsule should be maintained on that manufacturer’s product and should not be switched to another manufacturer’s capsule. As a result, none of the 10 pharmacies any longer switched individual patients to the cheaper NRIM SP - even when the prescription did not specify the brand.
I want to draw attention to 4.56 of the decision: “The CMA considers that the switching that occurred towards NRIM’s Product was not at a scale that meant NRIM’s Product exerted a sufficient competitive constraint on the Focal Product such that it would be included in the relevant markets. If NRIM had been a sufficient constraint it would have gained a much larger market share given its significantly lower prices.” In my view, the decision leaves room for an alternative interpretation.
The CMA contacted ten pharmacy groups during the course of its investigation covering approximately 50% of pharmacies in the UK and accounting for over 75% of NRIM’s total sales. As pharmacies have an incentive to dispense the cheapest medicine available, two out of the ten pharmacy groups contacted switched to dispensing NRIM PS to individual patients when it was launched and when there was an opportunity (i.e. the GP’s prescription was “open” and did not specify the brand). The other eight pharmacy groups informed the CMA that, in the period April to November 2013, they followed the principle of Continuity of Supply, rather than commercial incentives, when determining which phenytoin sodium capsule product to dispense. As a result, by November 2013 NRIM was dispensing an estimated 20-30% of all 100mg SP capsules in the UK (4.55 of the decision).
According to the decision, eight out of the ten pharmacy groups contacted were sufficiently concerned by the risk of therapeutic failure that they did not view the Focal Product and NRIM's Product as substitutes. In my view, this conclusion is less than convincing on the basis of the evidence presented. Of course, pharmacists should exercise caution when switching to a new product. That not all pharmacists would switch to a new brand immediately is only logical. It would not be surprising to me to hear that the largest pharmacy groups, probably most knowledgeable on regulatory matters, would come to a different estimation of the relative costs and benefits of switching brands when possible. I will explain in the next section that this appears to have been the case.
According to the decision, the two pharmacies that switched when possible accounted for 75% of NRIM’s total sales, i.e. these two groups supplied about 15-22.5% of the NRIM SPs and X% of Pfizer-Flynn SPs to the UK market. According to the decision, for the first eight months of 2012, 62% of prescriptions for phenytoin sodium capsules in England allowed for switching. Assuming that this percentages stayed the same over the next six months, one may assume that the two groups supplied about X = (15-22.5) * 38/62 = 9.1-13.8% of the of Pfizer-Flynn SPs to the UK market. In other words, these two groups that switched when possible accounted for 24.2-36.3% of the supply of SP in the UK. The average market share of the two groups that switched was thus about 12-18%. The average market share of the other eight groups was therefore at maximum 8-9%, but probably closer to 4-5% as the ten groups together only accounted for approximately 50% of pharmacies in the UK. That is, the average market share of the group that switched is significantly larger than the average market share of the group that stayed passive following the entry by NRIM.
This is reason to suspect that the two groups that switched could more in general be spearheading developments in the sector. It does not appear implausible that the smaller eight pharmacy groups would have switched if the two groups that switched would have reported positive results (i.e. economic benefits and no adverse effects on patients). One may therefore question whether the period assessed (April – November) was sufficiently long to conclude that eight out of the ten pharmacy groups contacted were sufficiently concerned by the risk of therapeutic failure that they did not view the Pfizer-Flynn SP capsules and NRIM's SP capsules as substitutes.
If a substantial number of the eight passive pharmacy groups would have switched after, say, the release of the annual accounts of the two switching groups, the conclusion on the market definition for at least the period April-November 2013 may well have been different. One may refer to the 2004 OFT guidance on market definition (3.6): “Customers may take time to respond to a sustained rise in the price of the focal product. As a rough rule of thumb, if substitution would take longer than one year, the products to which customers eventually switched would not be included in the same market as the focal product. Products to which customers would switch within a year without incurring significant switching costs are more likely to be included in the relevant market. However, the relevant time period in which to assess switching behaviour may be significantly shorter than one year: for example, in industries where transactions are made very frequently. A case by case analysis of switching is therefore appropriate.” The decision does not contain evidence of switching behaviour by pharmacy groups in general.
As the stated duration of the infringement was from September 2012 to date of the decision, the question of market definition is not moot.
Abuse of dominance in the Dutch rail sector
Dr. Rob van der Laan, 4 July 2017
The Dutch competition authority ACM recently published a (relatively rare) decision establishing the abuse of a dominant position. The ACM concluded that the incumbent rail services provider abused its dominant position by (primarily) submitting a loss-making bid for a local concession that could not be met by equally-efficient competitors without also making a loss. Whereas the local consession only related to a minute part of the Dutch railroads, the tender was a pilot project to assess the scope for further decentralization and regional competition. The loss of the concession could potentially result in the opening-up of larger parts of the Dutch rail network for competition and conversely winning of the regional concession by the incumbent could possibly prevent the emergence of competition in other regions. The incumbent was fined 40.950.000 Euro.
The analysis centers on the as-effective competitor test. The economic analysis appears to have been applied to the company at large, in effect a monopolist on most of the Dutch rail network. The analysis makes a distinction between the expected profit estimated on the basis of the winning of the regional concession as such (there appears to have been no dispute that the profit for the regional concession was negative) and the expected profit (avoided losses) of winning the regional concession in a wider context. As mentioned, this wider context was to prevent competition across a large part of its services: The incumbent internally used the metaphor of a house of cards – the loss of the single regional concession concerned could (or would) have repercussion in relation to other services and other regions. The incumbent appears to have considered its bid profitable when these wider effects were taken into account, however, according to the ACM, the bid was based on unsubstantiated expectations of growth of demand and the exclusion of risks. These expectations and exclusions had been the subject of internal discussion within the incumbent provider.
The ACM imposed a fine of €39m for the abuse of the dominant position sic. This was increased 5% (€1,95m) for a lack of cooperation with the investigation.
Dr. Rob van der Laan, 7 June 2016
De Autoriteit Consument en Markt voert de komende weken campagne om mensen alert te maken op kartelafspraken. Naar mijn mening een zeer lovenswaardig doel waar de economie netto profijt van kan hebben.
In het ACM-persbericht van 7 juni 2016 staat te lezen:
“Er zijn altijd mensen die weten van verboden afspraken tussen concurrerende bedrijven. (…) Een kartelafspraak is mensenwerk. Voorbeelden uit recente zaken:
in e-mails stond: ‘Streven moet zijn om de tarieven te verhogen’.
Of: ‘Een verhoging wordt door beide bedrijven noodzakelijk geacht’.
Een citaat van een manager in een bespreking tussen concurrerende bedrijven spreekt voor zich: ‘Our competitors are our friends, our customers are our enemies’.
Er zijn dus altijd mensen die weten van de verboden afspraken.”
Bij elk van de gegeven voorbeelden kan ik me binnen een bepaalde context een verklaring voorstellen die plausibel is en die niet duidt op een kartelafspraak. Ik ben niet nagegaan of de citaten vermeld staan in ACM-beslissingen, de bronnen worden niet vermeld in het persbericht. Daarom is het jammer dat tussen de bullet points en de conclusie een cruciale zin ontbreekt:
“Na zorgvuldig onderzoek door de ACM naar de context waarin de uitspraak is gedaan is vast komen te staan dat deze uitspraken betrekking moeten hebben gehad op verboden afspraken.”
Zonder deze toevoeging is het gevaar dat compliance officers meer van hun schaarse tijd zullen gaan besteden aan het redigeren van interne teksten in plaats van het voorkomen (en indien nodig beëindigen) van kartels – wat zeker niet goed zal zijn voor de economie.