Eurotunnel: when buying assets is a merger

When is an asset acquisition a merger? As the Eurotunnel litigation shows, the answer is not clear-cut.

The background is the 2011 liquidation of the cross-channel ferry company SeaFrance. It could not be sold as a going concern, so instead there was an asset sale. Eurotunnel bought three ferries and various other assets including the SeaFrance logos, brand and trade name, computer software, websites and domain names, and IT systems.

The Enterprise Act 2002 provides (in summary) that there is a relevant merger situation if one business acquires, “the activities, or part of the activities, of [another] business.” At the risk of stating the obvious, the focus should therefore be on whether the acquiring enterprise has acquired all or part of the activities of its target.

The OFT decided that Eurotunnel’s acquisition of the SeaFrance assets was a merger and should be prohibited. That decision was the focus of the first Eurotunnel case ([2013] CAT 30, Eurotunnel I”). The CAT held that, to decide whether Eurotunnel had acquired all or part of SeaFrance’s activities, the OFT should have asked three questions: (a) whether Eurotunnel had obtained more than “bare assets”; (b) if so, whether that placed Eurotunnel in a different position than if it had simply gone into the market and acquired the assets; and (c) if so, whether this difference turned what would otherwise be an acquisition of bare assets into an acquisition of the “activities of a business”.

This three-stage approach might seem a rather elaborate way of answering the simple question of whether Eurotunnel acquired some of SeaFrance’s business activities. However, Eurotunnel I was not appealed, and so that is the approach which the OFT (now the CMA) had to adopt when the matter was remitted.

One thing which becomes apparent, on reading the second Eurotunnel case (Eurotunnel II, [2015] CAT 1), is that the three-stage approach endorsed in Eurotunnel I leads to a range of questions about such things as the nature of a “bare asset”, or what kinds of assets Eurotunnel could have purchased on the market, which add a level of complexity to the analysis which is not obvious from the statutory language.

The key issue – which is really the third of the Eurotunnel I questions – is whether what was purchased was the activities of a business. On this key issue, the CAT in Eurotunnel II held that it is “a question of fact and degree” which “calls for the exercise of judgment by the decision-maker for which there is not necessarily a clear-cut answer” (at [74]).

Applying that approach, the CAT upheld the CMA’s decision that Eurotunnel had acquired part of SeaFrance’s business activities. Of particular importance was the fact that the combination of assets purchased enabled Eurotunnel to establish ferry operations more quickly, easily, cheaply and with less risk than had they been purchased on the market. Although it did not purchase the SeaFrance assets as a going concern, the reality was that it obtained much of the benefit of so acquiring them.

One issue which may benefit from further clarification in light of Eurotunnel II is whether it is ever appropriate to treat assets with the potential to be turned into a business activity as though they are already a business activity. The CMA placed some emphasis on the fact that Eurotunnel could, and indeed did, use the acquired assets to resume the business activities previously undertaken by SeaFrance within a very short period of time following the acquisition. That may be an important point, but of course the CAT in Eurotunnel I had held that the fact that (after the acquisition) Eurotunnel resumed SeaFrance’s business activities was not by itself sufficient to make the acquisition a merger.

The line between (on the one hand) the acquisition of a business activity, and (on the other hand) the acquisition of assets with the potential to create a business activity, therefore remains open for debate.

Leave a comment

Filed under Mergers

The costs of intervening

There is an interesting little point on costs buried away in last week’s decision in the “Ethernet” disputes in the Competition Appeal Tribunal (see BT plc v Cable & Wireless Worldwide Plc and others [2014] CAT 20).

Parties which intervene in CAT proceedings generally know that they are unlikely to recover their costs, even if they intervene in support of the party which is ultimately successful. There are, however, various exceptions to that principle – – and, indeed, in the Ethernet case itself some of the intervenors recovered some of their costs from the unsuccessful party.

But what is particularly unusual about the Ethernet decision is that an order was made requiring an intervenor to pay some of the successful party’s costs. As far as I am aware, that is the first time the CAT has made such an order.

The Ethernet disputes involved several appeals. BT intervened in Ofcom’s favour in the appeals brought by communications providers (Cable & Wireless and various others). Relatively early on in the proceedings, Ofcom effectively abandoned its defence on one of the issues in the appeals. BT, however, continued to defend Ofcom’s initial decision. The Tribunal said, at para 25:

“If and to the extent that Ofcom does not seek to resist an appeal but leaves the issue to be contested between the parties to the underlying dispute, we consider that the party in whose favour Ofcom determined the dispute and then intervenes in support of Ofcom’s decision will in practice perform the role of respondent in the appeal. In those circumstances, it may well be just for that intervener to be treated as regards costs in the same way as a private party to an appeal.”

BT was ordered to pay 75% of the appellants’ costs after the date on which Ofcom made it clear it was not defending the reasoning in its decision.

These are relatively unusual facts, but they do raise a further note of caution for intervening parties.

Leave a comment

Filed under Procedure, Telecoms

Applying interest in damages claims

The Competition Bulletin is pleased to welcome the latest in our series of blogs by Oxera Consulting on key economic concepts for competition lawyers. In this blog, Enno Eilts, a Senior Consultant, discusses issues connected with the calculation of interest in damages actions.


Compensation for antitrust and commercial damages is often claimed after a significant delay, sometimes many years after the damage has been suffered. Interest is routinely added to such claims to account for the effects of money losing its value over time, as well as the lost opportunity to the injured party from having the capital at its disposal.

While the recently adopted European Directive on antitrust damages actions will help to iron out some cross-jurisdictional differences (for example, with regard to disclosure and limitation periods), this applies to a lesser extent to the question of applying interest.[1] Here, member states are required to lay down the rules. This makes the application of interest an important consideration for claimants—and defendants—when choosing the forum in which to bring a claim.

So how should the interest component of a damages claim be calculated? And what difference does it make?

This blog post illustrates the importance of the interest component in damages claims, and compares how interest might be applied in the UK, Germany and the Netherlands—the three jurisdictions that currently seem to attract the largest number of antitrust follow-on claims. It then discusses some of the relevant economic considerations when applying interest.

What difference does it make?

The law on applying interest in damages cases is complex and varies between countries. Apart from the period over which interest is to be applied, the central questions relate to the rate that is used (i.e. a statutory or an economic rate) and the methodology applied (simple versus compound interest).

The table below illustrates the difference that using different rates and methodologies for calculating interest can make to the value of a claim in the UK, Germany and the Netherlands. In each example, the claimant incurred a damage of £10,000,000 spread equally over the years 1999 to 2003.


Calculating interest in the UK, Germany and the Netherlands—illustrative examples using typically applied interest rates


UK Germany Netherlands
Value of the claim before adding interest £10,000,000 €10,000,000 €10,000,000
Interest rate used Bank of England base rate +2%[2] Bundesbank base rate +5%[3] European Central Bank base rate +7%[4]
Methodology applied Simple Simple Compound
Value of the claim in 2014 after adding interest £16,600,000 €18,400,000 €31,900,000


Note: The increase in the claim value is calculated on the basis of a damage that occurred between 1999 and 2003 (in equal proportions each year) and is claimed in 2014. The legal questions around calculating interest are complex. The purpose of the above table is to illustrate the effects of using different interest rates and methodologies on the value of a claim, rather than to make the case for or against the use of these rates and methodologies.

Source: Oxera.


The table illustrates that, in the case of damages that occurred between 10 and 15 years ago, the interest component of a claim brought in the UK or Germany in 2014 could increase the value of the claim by 66% or 84%, respectively if simple interest is used. In contrast, applying interest on a compound basis in the Netherlands could more than triple the value of the claim.

Why add interest?

As victims of antitrust and commercial damages tend to be unaware of having been harmed, they often claim compensation only years after the harm was suffered. For example, in the case of the car glass cartel, about a decade elapsed between the start of the cartel period in 1998 and the time when potential victims are likely to have found out that they might have been harmed—i.e. when the European Commission published its decision in 2008.[5]

From an economics perspective, adding interest to a damage amount that was incurred in the past is sensible. This is because £1 today has a greater value than £1 in ten years’ time. Damages awarded ten years ago would therefore have had a higher value to the victim at the time than if they had had to wait until today to be awarded the damage. This is partly because of the loss of value over time (due to inflation), but it is also due to the loss of opportunities (for actually using the money).

Consistent with economic principles, compensation rules in EU law mean that damages awards should include interest. The objective is to place the injured party in the position it would have been in had there been no infringement.[6] This seems to imply not only that the direct damage from the competition law infringement should be reflected when determining the damages awarded to the claimant, but also that compensation for the loss of opportunities due to the deprival of funds should be considered.

However, legal rules and practices vary significantly across jurisdictions, and across cases within jurisdictions. Some approaches to applying interest are somewhat at odds with economic principles—if the objective is, indeed, adequate compensation. One specific issue is whether interest is calculated on a simple or a compound basis. Another is that various jurisdictions stipulate the use of statutory interest rates rather than commercial/market-based interest rates.

What interest rate should be applied?

If the objective is to place an injured party in the position it would have been in had there been no infringement, the relevant question in terms of adding interest is what the party would have done with the funds that it lost at the time. One metric that takes this aspect into account is the cost of capital.

The cost of capital reflects the ‘normal’ returns that a claimant can be expected to earn in the long run. Thus, damages uprated at the cost of capital would capture the expected returns that the claimant could have earned on the amounts lost had they been available for investment.

An alternative option is the risk-free rate. This is usually approximated by the rate on a virtually risk-free investment such as a government bond. The rationale for this is that the repayment of damages is certain once awarded (subject to the defendant’s inability to pay). Using the risk-free rate therefore only compensates the claimant for the time value of money, without a risk component.

As illustrated in the table above, national courts have tended to adopt a ‘middle way’ between the cost of capital and the risk-free rate by applying statutory or quasi-statutory rates of interest. These tend to be based on a central bank’s base rate plus a premium. Using a statutory interest rate is often seen as ‘likely to do rough justice in most cases’ while avoiding the complication of enquiring into the particular financial circumstances of an individual victim.[7]

In principle, a company’s cost of capital can be higher or lower than the statutory interest rate applied by the court, depending on the relevant jurisdiction. Which rate results in a higher or lower value of claim therefore depends on the circumstances of the case and the jurisdiction in which the claim is brought. The risk-free rate, in contrast, is lower than the cost of capital and is also likely to be lower than the statutory interest rate, as the latter usually incorporates a mark-up in excess of the risk-free rate (e.g. defined as base rate + x%, as shown in the table).

Compound or simple?

Interest rates can be applied as simple and compound interest. When applying the simple methodology, interest is calculated solely as a percentage of the principal sum. When the interest is compounded, the calculation includes interest on accumulated interest from prior periods. As such, adding compound interest increases the claim value by a larger degree than adding simple interest.

From an economics perspective, compounding interest is the usual, and conceptually correct, approach. A savings bank, for example, pays interest on a whole balance, including past interest. Nonetheless, there are many instances where the legal framework requires simple interest to be applied. For example, certain provisions in German law explicitly states that interest should be calculated on a simple basis.[8]

However, there does seem to be appetite for change. For example, ten years ago the Law Commission for England and Wales recommended the application of compound interest, as it better reflects commercial reality.[9] In the Sempra Metals case, the UK House of Lords ruling also notes that a claim based on simple interest ‘will inevitably fall short of its true value’.[10]

Interest calculations in damages cases brought in the Netherlands tend to be more in tune with economic principles. Here, interest on past damages is routinely calculated on a compound basis (as illustrated in the table).

Concluding remarks

Courts generally accept that, to adequately compensate victims, interest should be added to damages suffered in the past. However, the rates used and the methodologies applied to calculate the interest component of a claim vary greatly across jurisdictions. This can have a significant impact on a claim’s value.

While some cross-jurisdictional differences will be reduced by the Directive on antitrust damages actions, a significant degree of flexibility is likely to remain with regard to applying interest. This makes the interest component an important aspect of a claimant’s choice of jurisdiction for bringing a claim.


[1] Directive of the European Parliament and of the Council on certain rules governing actions for damages under national law for infringements of the competition law provisions of the Member States and of the European Union (November 2014).

[2] Case: 1178/5/7/11 2 Travel Group PLC (in liquidation) v Cardiff City Transport Services Limited.

[3] Pursuant to §288 German Civil Code (Bürgerliches Gesetzbuch, BGB).

[4] Article 6:119a Dutch Civil Code (Burgerlijk Wetboek).

[5] European Commission (2008), ‘Summary of Commission Decision of 12 November 2008 relating to a proceeding under Article 81 of the Treaty establishing the European Community and Article 53 of the EEA Agreement (Case COMP/39.125 — Car glass)’, available here.

[6] Joined cases C-295/04 to C-298/04 Manfredi [2006] ECR I-6619, 95.

[7] Credit Lyonnais SA v Russell Jones & Walker [2003] PNLR 2, Laddie J at 27-28.

[8] § 289 Bürgerliches Gesetzbuch.

[9] Law Commission, Pre-Judgement Interest on Debts and Damages, Report Law Com No 287, 24 February 2004.

[10] Sempra Metals Ltd v. Revenue & Anor [2007] UKHL 34, 18 July 2007.

Leave a comment

Filed under Abuse, Agreements, Damages, Economics

What’s the plot? Conspiracy and 19th Century comic opera (again)

Ever since Johnson v Moreton [1980] AC 37 (61E-G per Lord Hailsham: ‘we should have to adopt the carefree attitude of the Mikado…’), references to Gilbert and Sullivan have been gaining ground in the judgments of our higher Courts. When last year Arden LJ rejected the argument, advanced by the claimant victim of a cartel, that it suffices to establish the intention requirement for the tort of unlawful means conspiracy that the claimant forms part of a class of persons against whom a cartelist’s wrongful acts were targeted, she did so by reference to The Gondoliers:

‘it deprives the requirement of intent to injure of any substantial content. It is tantamount to saying it is sufficient that the conspirators must have intended to injure anyone who might suffer loss from their agreement. If I might say so, the submission is reminiscent of the circularity of words in the Gondoliers that “when everyone is somebody, then nobody is anybody”’.

(See W.H. Newson [2013] EWCA Civ 1377 at [41], blogged here by Andrew Scott).

However, in a recent judgment in the Air Cargo litigation, the proposition for which Arden LJ invoked The Gondoliers has encountered judicial disapproval: Emerald Supplies Ltd v British Airways plc and others [2014] EWHC 3514 (Ch) (not to be confused with the judgment on confidentiality issues in the same case handed down on the same day, the subject of this blog post by Eesvan Krishnan). Refusing to accede to BA’s pre-disclosure strike-out application of the air cargo operators’ claim in unlawful means conspiracy, Peter Smith J confessed that he found Arden LJ’s reasoning difficult to follow, perhaps for ‘my lack of knowledge of Gilbert & Sullivan operettas (save the song “I am an Englishman”)’: [76]-[77].

Peter Smith J could not see how it would deprive the requirement of intent to injure of substantial content to accept the “targeted class of persons” argument that had been advanced in Newson, and considered Arden LJ’s remarks an unnecessary gloss on the principles established in OBG v Allan [2007] UKHL 21: see his judgment at [78]. Newson, he suggested, could be distinguished on its facts, and in any event he considered Arden LJ’s observations to be strictly obiter: [80]-[81].

It is unclear whether the judge’s reference to HMS Pinafore (‘in spite of all temptations/ To belong to other nations,/ He remains an Englishman!’) is intended as a nod to a Eurosceptic trend among our senior judiciary. But if Newson seemed to raise the bar for competition litigants running good old common law claims, Emerald appears to lower it – and marginally to increase the chances of a future judgment of the High Court featuring the following exchange between The Grand Duke’s Julia and Ludwig:

‘JUL. Bah! Lots of people eat sausage-rolls who are not conspirators.
LUD. Then they shouldn’t. It’s bad form. It’s not the game.’

Leave a comment

Filed under Agreements, Damages

“It’s too late baby, now it’s too late”: limitation, competition claims and knowledge

How much knowledge does a potential claimant need before time begins to run against a competition claim against a party alleged to have breached competition law? This was the key question addressed by Mr Justice Simon in the first case in which an English Court has had to consider the effect of s.32 of the Limitation Act 1980 (“LA”) in the context of a competition claim.

The judgment is the latest instalment of the MIFs (“multi-lateral interchange fees”) saga (see, e.g. Tom Coates’ blog on the CJEU’s recent decision in the MasterCard case).

In Arcadia Group Brands Limited and others v Visa Inc and others [2014] EWHC 3561 (Comm) claims were brought by a number of major UK retailers such as Argos and B&Q against various Visa entities (e.g. Visa International and Visa Europe). The claimants alleged that Visa had charged anti-competitive MIFs since 1977, and sought over £1 billion in damages.

The defendants applied to strike out the claim for the period between 1977 and 2007 on the basis that the claimants were out of time to sue for that period. This would effectively reduce the claim by approximately £500 million (as noted at [20]). The retailers relied upon s.32 LA to argue that Visa had concealed relevant facts and that they could therefore not have sued before 2007. Indeed, their counsel argued that the limitation clock had not even started to run given that not all the relevant facts were known at the time of the hearing (at [26]).

Simon J identified the conflicting interests at issue: on the one hand, the “important public interest that claimants should not be prejudiced where they lack sufficient information to advance a claim”, on the other “the general policy of limitation legislation, the public interest in ensuring certainty and finality in litigation” (at [107]). After a clear exposition of the legal principles, the Judge found for the defendants, considering that the “[f]acts which are still unknown and are not essential to complete the cause of action cannot amount to relevant facts for the purpose of s.32(1)(b)” (at [28]). In particular, he concluded that “the trigger for the running of time for limitation purposes is not the discovery of every potentially relevant fact in the broadest sense”. He then pointed to five European Commission documents (press releases, a negative clearance decision, an exemption decision and an Article 19(3) Notice) (at [52]-[69]), three OFT documents (press releases and decisions relating to MasterCard’s MFIs) (at [76]-[86]) and the European Commission’s 2007 MasterCard Decision to accept the defendants’ case that “the Particulars of Claim derived from material which was available before” the relevant dates six years before the claims were brought.

The judge noted (at [33]-[34]) the Court of Appeal’s decision in KME Yorkshire Ltd and ors v Toshiba Carrier Ltd (UK) and ors [2012] EWCA Civ 1190 (as to which, see Tom Richards’ post here), which had recognised a ‘generous approach’ towards claimants’ pleadings when applications are made to strike out competition claims. However, he rejected a contention that a claim for breach of Article 101(1) TFEU was particularly complex or fell within a developing area of law, concluding that “competition cases […] do not fall within an exceptional category calling for a different approach to the application of s.32(1)(b)” (at [38]).

In sum, although accepting that “the full picture was not available” (at [101]), the Judge felt that “the Claimants are focussing on matters about which they might want reassurances before bringing a claim, but which do not constitute matters which are essential to pleading it” (at [103]). He emphasised that this was “not a case of a ‘secret cartel’ operating over many years without the knowledge of victims and the authorities” but one which was a matter “of public knowledge, which had been notified to the competition authorities” (at [106]).

This is now the leading case on the application of s. 31(1)(b) LA in competition damages cases. In some ways, it acts as a counterbalance to the alleged ‘special pleading’ rules applicable to competition claims, placing a considerable onus on potential claimants to be proactive when the whiff of a potential infringement of competition rules enters the public domain.

Indeed, this is an interesting result given that the claimants candidly accepted that “[t]he trigger for the proceedings was […] the dismissal of the appeal against the 2007 MasterCard Decision by the General Court in May 2012” (at [105]). This was, apparently, too late. Such a finding reinforces the sentiment that the English approach to the Masterfoods line of case-law is correct in encouraging national proceedings to go ahead as far as possible before EU proceedings are concluded.

Finally, the Court was well aware of the backdrop to the issue of limitation in the context of private damages claims. The Claimants pointed to the draft Article 10 of the new Directive on such claims (the corrected version of which the Council is expected to adopt soon). This provides that national limitation periods must (i) be at least 5 years (ii) shall not begin to run until the infringement of competition law has ceased and (iii) are triggered only when the claimant “knows or can reasonably be expected to know” of (a) the behaviour and the fact that it infringes competition law (b) the harm he has been caused by that behaviour and (c) the identity of the infringer. The Judge emphasised, however, that this was not the present state of the law and so would not affect his reasoning. Nevertheless, in the future the approach may well be more generous to claimants, particularly in cartel cases where the infringement is often ‘secret’. But even in such cases, if any public information identifies key facts, potential claimants are put on notice. Claimants will need to be proactive and defendants vigilant to challenge any perceived delays in initiating litigation.

Leave a comment

Filed under Abuse, Agreements, Damages, Procedure

High Court tests the limits of confidentiality in EC infringement decisions

The European Commission came in for some stern criticism from the High Court this week, in a case which looks set to test the boundaries of confidentiality in EC infringement decisions: see Emerald Supplies v BA [2014] EWHC 3515 (Ch).

The background is the 2010 EC decision fining BA and eleven other airlines a total of €800m for operating a global cartel for air freight services. Hundreds of claimants are seeking damages, and they sought disclosure of a copy of the decision – which, remarkably, has not yet been made public by the EC.

The problem is that the decision apparently contains comments alluding to potential wrongdoing of various airlines which those airlines were unable to appeal – – either because they were not addressees of the decision, or because the offending passages were not essential to the operative part of the decision and therefore could not be challenged on appeal. Relying on the case of Pergan Hilfsstoffe Fur Industrielle Prozesse GmbH v Commission [2007] ECR II-4225, those airlines argued that disclosing such references would breach the EU “presumption of innocence”, and that it should not be permitted.

Whether Pergan does require all such references to be withheld is a complex question. The judge in Emerald (Peter Smith J) was initially prepared to accept that Pergan does impose such a requirement. Eventually, however, exasperated by the practical difficulties such a rule would create, he changed his mind.

The judge’s first order (made in April) was for affected parties to agree on redactions to protect the so-called “Pergan rights”. This resulted in a document so comprehensively redacted that it was useless to the claimants. The claimants then suggested that the judge himself carry out the redactions necessary to protect confidentiality, but he considered the task to be simply impossible in the circumstances. Instead, he ordered the disclosure of the decision – unredacted except for ‘leniency materials’ and material subject to legal professional privilege – to a confidentiality ring comprising the claimants and the Part 20 defendants (a group which includes various airlines, including non-addressee airlines who have not seen the decision).

The disclosure will be subject only to the condition that the claimants were barred from using the decision to commence proceedings without the permission of the court. That restraint, the judge thought, allays all legitimate Pergan concerns: namely, that the decision could contain observations or findings that any airline had not had an opportunity to deal with; that the decision contained material that would cause damage if it went into the public domain; and that the decision might identify other people against whom claims could be brought. Through a confidentiality ring subject to that condition, the judge thought, all the parties – the claimants, BA, and the Part 20 defendants alike – would enjoy an equality of arms in prosecuting and defending the action.

Complaining that the EC had failed to comply with the duty of sincere co-operation between the Union and Member States (Article 4(3) of the TEU), the judge rejected the submission that his approach would rob any future decision on redaction by the EC of any practical effect – placing great emphasis on a letter by the EC to the Court in which the Commission appeared to be content for the Court to decide the question.

These types of issues will arise in many cases, including cases where the EC has published a public version of the decision (since there will still be a question as to whether addressees of the decision should be required to disclose the confidential version of the decision in the context of a private damages claim). However, the fact that there was no public decision in this case undoubtedly made the judge’s job more difficult. It is difficult to disagree with his criticisms of the EC’s “one speed molasses like approach” to redacting its decisions, which he said was “completely unacceptable” (paragraph 3 and 27). The appropriateness of the Judge’s decision will be for the Court of Appeal to decide in the coming months: recognising the importance of his decision, he has granted permission to appeal.

1 Comment

Filed under Abuse, Agreements, Damages, Procedure

Competition law and public services: insights from the OFT report into higher education

Recent public sector reforms have relied on choice and competition to increase the quality and quantity of service provision, whilst also controlling cost, through a programme known as Open Public Services. The use of choice and competition gives rise to public service markets, and ensuring that these markets function effectively is one of the Competition and Markets Authority’s declared objectives. Higher education constitutes one of the larger public service markets, and to understand how the market for undergraduate education in England functions, in October 2013, the OFT launched a Call For Information. Amongst other things, the OFT wished to consider whether it was plausible for universities to have arrived at a uniform fee for all their undergraduate courses without colluding, and whether the way prospective undergraduates apply for university places could harm competition between institutions, to the detriment of students. The OFT’s higher education report, published in March 2014, provides useful insights into the role of competition law in public service markets and of the challenges of applying competition law in public service markets.

Competition Advocacy

One of the OFT’s clearest findings was that the regulatory environment, rather than service providers, may limit choice and prevent competition. There is also a perceived tension between cooperation and competition, and a more general debate over the desirability of choice and competition in the public service context. This gives an indication of just how much work the CMA, using competition advocacy powers available under section 7 of the Enterprise Act 2002, has to do in order to explain the extent to which choice and competition are an appropriate framework in which to provide public services.

The CMA has moved a long way from the OFT’s initial view, set out in its BetterCare decision and Policy note 1/2004, of competition law’s limited relevance to public service markets. Since 2004 a number of documents (for example: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, and 16) indicate just how keen the CMA and its predecessor have been to show that competition can and has been used to create incentives to improve quality; reduce cost; or innovative in the provision of public services. What we have yet to see is enforcement action.

The substantive approach

The OFT considered whether the way prospective undergraduates apply for university places using the centralised admission system, UCAS, could harm competition between institutions, to the detriment of students. Two features of the application process are that students are permitted to apply for no more that five courses in an application cycle and are prohibited from applying to both the University of Oxford and the University of Cambridge in any given application cycle (para. 6.24-6.30).

Both measures can be described as restrictions of choice. There is an idea that the fundamental purpose of competition law is to protect consumer choice, so that the examined measures would immediately be suspect. The importance of choice has also been emphasised by the CAT (see, for example, paras 255, 468 and 585 of Genzyme v OFT [2004] CAT 4). The OFT concluded that it is “unlikely” that the restrictions on choice “would result in a risk that either institution could manipulate price, quality or choice, to the detriment of students.” (para. 6.27). The OFT arrived at this conclusion by considering competition in two stages: pre-application (when applicants are choosing to which institutions they would like to apply), and post-application (once applicants have submitted their UCAS forms). The OFT report considers the freedom to choose which institutions to apply to drives the competitive response on the supply-side so that competition is most important in the pre-application phase:

“the main factors on which higher education institutions compete to attract students – course features, institutional facilities and tuition fees, for example – are generally set in the pre-application period” (para. 6.27).

Pre-application there is no restriction on choice or competition, since: “Applicants are free to apply for any course offered across all UK higher education institutions, and institutions compete to attract students.” (6.34).

The OFT’s benign treatment of the Oxbridge restriction raises the question of whether it is open to other institutions to adopt similar restrictions. For example, would it be compatible with competition law for LSE, UCL, KCL, Imperial, and QMUL to enter into an agreement whereby a student may apply only to one of the institutions? The OFT analysis suggests that it would be entirely compatible with competition law because there is plenty of pre-application competition and post-application there remains competition from the four other institutions to which the student has applied (6.27). Even if restrictive, the OFT report suggests that the agreement would be justified if it enabled “a more in-depth assessment of each candidate” (6.29). This does not seem to be the correct answer, but the OFT report does not indicate why it is incorrect, or how to distinguish the Oxbridge limitation from this hypothetical case.

Since competition is not for applications but for acceptances, it would seem sensible to consider three phases in the process: in phase I the student makes an application; in phase II universities issue offers to applicants; in phase III the student compares offers and decides which, if any, to accept. The institution’s chance of success in phase III is determined by the offer it makes in phase II. Further, the fewer offers a student receives in phase II the greater the chance of an institution’s offer being accepted in phase III. The potential dampening of competitive fervour that might result from both measures can then be seen from a letter published in the THES to justify the prohibition on applying to both Oxford and Cambridge, claiming:

“If a significant proportion of the applicants to whom [Oxford] offered places were liable to go instead to Cambridge, then to avoid lots of places going to waste, we would have to treat admissions as a central university process, playing the statistics of large numbers rather than selecting the students for our own colleges.”

The justification I see is that a student may not apply to both Oxford and Cambridge because in phase III an offer of a place at Oxford may be spurned in preference for a place at Cambridge. In order to ensure that the offer were not rebuffed the institution would need to increase the attractiveness of its offer or accept applicants may decide to take up a more attractive offer. Perhaps the student would be swayed to accept the offer if it included a scholarship or other discount on the cost of completing the course; or if the fees were generally lower; or because the offer included a guarantee of accommodation for the duration of the course—i.e., competition on things that matter, over and above that which occurred in phase I.

There is both US experience of competition law being applied to similar measures, and a debate as to whether the competition that exists is of the wrong sort (1, and 2). The OFT consider that even if the measures do restrict competition the restriction may be justified on the grounds that “since each additional choice that an applicant makes puts a cost on the institution, it may be efficient to restrict the number of choices that each applicant can make” (6.35), particularly since this may enable “a more in-depth assessment of each candidate” (6.29). The number of applications to Oxford has increased by 46% in the past 10 years. Cambridge has also seen an increase. It is thus not clear that the restrictions are necessary to limit the number of applications, particularly since there is no absolute limit on the number of applications institutions can receive and institutions are actively engaged in attracting applicants.

Further, even in public service markets, any restriction must be “to the benefit of students” or users (6.35). The identified benefits would appear to accrue to the institutions rather than to the students. The OFT held over 50 meetings and analysed over 80 submissions (1.5; and 2.12). Who the OFT met and the content of the written submission is not intended to be disclosed, the OFT stating: “For reasons of commercial confidentiality, material may appear in an anonymous, aggregated, or otherwise redacted form” (para. 2.9). This prevents the veracity and strength of the submissions being tested or countered through public debate.

The OFT approach may be compared to antitrust action currently playing out in the US—CollegeNet, Inc., v The Common Application, Inc., filed on 8 May, 2014. CollegeNet is challenging measures it considers to prevent or discourage applications to institutions other than through Common Application, to the detriment of students. This action may provide a useful review of the benefits of a single application system and of the restrictions necessary for such benefits to be obtained. In any event, the CMA response to the OFT report is eagerly awaited.

Leave a comment

Filed under Mergers