Introducing new EU Merger Control thresholds to tackle Big Data Concerns? : A comparative analysis of the jurisdictional merger control thresholds in place in Germany, the UK, Denmark, the Netherlands and Norway

(1)

The

Date of Submission: 25/07 Supervisor: Dr. Or. Brook Word count: 14284 (excl. abstract, table of content, tables and bibliography)

INTRODUCING NEW EU MERGER

CONTROL THRESHOLDS TO TACKLE BIG

DATA CONCERNS?

A comparative analysis of the jurisdictional merger control

thresholds in place in Germany, the UK, Denmark, the

Netherlands and Norway.

Sari Corrijn

University of Amsterdam – Master thesis LL.M. International and

European Law: European Competition Law and Regulation

(2)

Abstract

Mergers between companies that possess important Big Data can create significant advantages in the competitive process. The existing EU merger control system is based on turnover notification thresholds and might be inadequate to submit mergers raising anti-competitive concerns due to Big Data to the merger review process. This has led to the following research question: ‘Should the EU introduce new jurisdictional merger control

thresholds to tackle Big Data concerns, considering the need for legal certainty, a cost-effective merger control system, and a sufficient local nexus?’

This question is addressed by comparing five jurisdictions with different merger notification thresholds that are able to tackle mergers with Big Data concerns. Those five jurisdictions are Germany, upholding a ‘size of transaction’ test, the United Kingdom, keeping a voluntary notification system with a ‘share of supply’ test, Denmark and the Netherlands, both working with ‘sector-specific’ merger control thresholds, and Norway, maintaining a ‘flexible’ merger control system. The comparison is based on four normative benchmarks, i.e. the degree of tackling Big Data concerns, legal certainty, cost-effectiveness, and a sufficient local nexus.

The comparison has led to the following findings. The German ‘size of transaction’ threshold raises some capacity concerns, but is in general cost-effective. However, it lacks sufficient legal certainty and a clear local nexus, which could interfere the division of competences between the European Commission and the NCAs. Nevertheless, many of these issues can be resolved by an extensive clarification of ambiguous concepts.

The UK ‘share of supply’ test in a voluntary merger control system is rather cost-ineffective, due to the need for continuous monitoring of the market and substantial burdens placed on the merging parties. Additionally, it fails to ensure sufficient legal certainty, but the local nexus criterium is plainly satisfied.

‘Sector-specific’ merger control systems demand more capacity on part of the CA. Furthermore, ‘sector-specific’ turnover thresholds fulfill the requirements of legal certainty and guarantee a sufficient local nexus. However, the level of turnover for the specific thresholds necessitates a careful consideration to tackle both data concerns, while still maintaining a sufficient EU nexus.

Finally, the Norwegian ‘flexible’ system demands more capacity because of the need for constant surveillance. Furthermore, it fails to guarantee legal certainty and a sufficient local nexus, which could interfere with the division of competences between the European Commission and the NCAs. Guidelines as to when the discretionary power of the CA would be exercised are essential if such a system would be introduced in the EU.

Based on the comparison according to the normative benchmarks, the German ‘size of transaction’ test would be the best option for the EU to introduce as an additional merger control threshold to tackle Big Data concerns. However, an extensive clarification of the concepts used in the test will be required. Finally, this thesis proposes to introduce a combination of the ‘size of transaction’ test with ‘sector-specific’ thresholds to fully and directly address Big Data concerns in mergers.

(3)

Introducing new EU merger control thresholds to tackle big data concerns?

1. Introduction and methodology

1.1. A Big Data enforcement gap in the EU Merger control system?

Big Data is becoming a tremendous valuable economic input for businesses. As a central feature in the business process, Big Data offers companies significant advantages in the competitive process.1 Therefore, it cannot be ignored from a competition law perspective. The primary concerns, caused by the collection and processing of enormous volumes of data, lies in the risk of foreclosing competition.2

Accordingly, Big Data can be a source of market power in a given relevant market. The collection and possession of Big Data, generates market power in predominantly two instances: a) when the data has a significant role as an input in developing goods or carrying out services, and b) when competitors cannot reasonably collect, replicate, purchase or substitute the data.3

Consequently, mergers in data related markets can cause various risks. For instance, a merger could result in differentiated data access and increase the concentration of data in the market. The new entity can obtain important competitive advantages by combining different data sets. This could especially raise competition concerns if competitors are incapable to reproduce information extracted from it. When the merging parties already hold strong market positions in upstream or downstream markets, they could foreclose these markets for competitors, i.e. by leveraging market power as a result of data into other markets.4

The EUMR seeks to avoid mergers which significantly impede effective competition within the internal market.5 However, current EU rules on merger notifications may be inadequate to reach this objective when a concentration involves Big Data. Jurisdictional merger control thresholds aim to exclude from review transactions that have no anti-competitive impact in a given

1_{A. Fidelis, Z. Ortac, Data driven mergers: a call for further integration of dynamic effects into competition analysis, Master}

thesis, Barcelona GSE, 2017, p. 3 (hereafter: A. Fidelis, Data driven mergers); G. Colangelo, M. Maggiolino, ‘Big Data as misleading facilities’, European Competition Journal, 2017, Vol. 13 No. 2-3, p. 250.

2_{M. Ristaniemi, T.A. Puutio, ‘The power of data: is it worth all the hype?’, KLCB, 2016 (hereafter: M. Ristaniemi, ‘Power of}

data’).

3_{I. Graef, ‘Market definition and market power in data: the case of online platforms’, Kluwer Competition law, 2015, p.}

486-489; see also M. Ristaniemi, ‘Power of data’, supra fn. 2.

4_{A. Fidelis, Data driven mergers, supra fn. 1, p. 16.}

5_{Article 2 Council Regulation (EC) No139/2004 of 20 January 2004 on control of concentrations between undertakings,}

(5)

jurisdiction. Accordingly, the purpose is to identify transactions that may have an anti-competitive impact in a jurisdiction and, therefore, should be reviewed.

The current EU merger control system is based on a mandatory notification system, where the undertakings concerned are obliged to notify a proposed merger to the European Commission if the EUMR jurisdictional thresholds are exceeded. The existing jurisdictional thresholds are based on turnover.6 However, turnover can be inadequate to tackle Big Data concerns. This is demonstrated by the Facebook/WhatsApp merger,7 in which the concentration did not meet the turnover thresholds, despite considerable chance of leading to anti-competitive effects in the EU.

In the context of a growing digital economy and the increasing value of Big Data, mergers between companies which fail to generate high turnovers, can possibly raise anti-competitive concerns. For example, some digital services are offered free of charge. Consequently, the company’s revenues may escape the turnover thresholds. Accordingly, transactions can comprise target undertakings whose turnover or market share does not reflect their real competitive significance.8 Hence, potentially harmful mergers may evade merger notification and ex ante review by the EC and NCAs.9

However, the existence of the plausible enforcement gap is still under debate and remains to be proven by empirical evidence. It should be noted that the case referral system of Articles 4, 9 and 22 EUMR, is possibly sufficient to fill in the alleged enforcement gap. The referral system allows the Commission, the merging parties and the NCAs to refer from one or several NCAs to the Commission or vice versa. For instance, when it appears that a case can be better handled at EU or national level respectively.10 This was the case in the Facebook/WhatsApp merger,

6_{Article 1 EUMR.}

7_{Case COMP/M.7217, Facebook/WhatsApp v. Commission (2014).}

8_{T. Cowen, ‘Big Data as a competition issue: should the EU Commission’s approach be more careful?’, ENLR, 2016, p.}

14-23.

9_{M. Ristaniemi, ‘Power of data’, supra fn. 2.}

10_{European Commission (DG Competition), Consultation on Evaluation of procedural and jurisdictional aspects of EU}

merger control (hereafter: Public consultation merger control), 2017; U. Von Koppenfels, ‘A Fresh Look at the EU Merger Regulation? The European Commission’s White Paper “Towards More Effective EU Merger Control”’, Liverpool Law

(6)

where the case eventually did end up at the European Commission. Nevertheless, this thesis is built on the hypothesis that the enforcement gap is real.

1.2. Methodology and approach

The existing EUMR turnover thresholds may be inadequate to subject mergers involving Big Data to merger review. The European Commission should be able to carry out an analysis of anti-competitive effects of mergers between companies that possess important and/or large data sets, even if they do not generate high turnovers. Therefore, this thesis will answer the following research question:

‘Should the EU introduce new jurisdictional merger control thresholds to tackle Big Data concerns, considering the need for legal certainty, a cost-effective merger control system, a sufficient local nexus?’

The research question is answered by conducting a comparative legal research. Five national merger control systems, viz. Germany, the United Kingdom (UK), Denmark, the Netherlands, and Norway, are compared on a functional basis, according to four normative benchmarks. The comparison does not only take into account the law as such, but also considers case law, policy documents and literature. The five jurisdictions were chosen because they each have specific merger control tests, which are able to submit mergers involving Big Data to the merger review process.

Germany takes into account Data concerns by employing a ‘size of transaction’ test, where the value of Big Data is considered for the transaction value.

The UK utilizes a ‘share of supply’ test in a voluntary merger notification system. Since Big Data has significant competitive advantages, the possession of important data sets can increase a company’s share of supply in a given market. The UK ‘share of supply’ test is, therefore, able to tackle Big Data concerns.

(7)

Denmark and the Netherlands are assessed together, since they both hold ‘sector-specific’ merger control thresholds. Big Data concerns can be addressed by introducing specific thresholds in sectors where Big Data particularly leads to anti-competitive effects.

Finally, Norway maintains a ‘flexible’ merger control system, which allows the NCA to order notifications below the mandatory turnover notification thresholds. This permits the NCA to submit mergers involving Big Data to merger review, even if the mandatory thresholds are not exceeded. A more detailed explanation as to why these jurisdictions allow to tackle Big Data concerns, is given in chapter 2 of this thesis.

The comparison of the various jurisdictions is based on following four normative benchmarks: 1) degree of tackling Big Data concerns, 2) legal certainty, 3) cost-effectiveness of the merger control system, and 4) sufficient local nexus. These benchmarks are chosen because they are fundamental to shape a proper EU merger control system dealing with mergers involving Big Data. Accordingly, they create essential guiding principles in assessing the added value of new jurisdictional thresholds. Their fundamentality follows from the EUMR, the ICN Recommended Practices11 and the doctrine.

The thesis is structured as follows. Chapter 2 explains the substance of the jurisdictional tests in place in the five jurisdictions and their ability to tackle Big Data concerns. Chapter 3 contains the main comparison of the jurisdictions in four sections based on the four normative benchmarks, with a fifth final section synthesizing the analysis. Chapter 4 concludes whether the EU should introduce new merger control thresholds to tackle Big Data concerns, considering the analyzed benchmarks. Therefore, this chapter entails the main take-aways of the comparison, an answer on the research question, and finally a proposal of a new kind of thresholds to tackle Big Data concerns.

11_{ICN Recommended Practices for Merger Notification and Review Procedures with Working Group Comments as amended}

(8)

2. How do the national merger control systems tackle Big Data concerns?

In order to assess whether new jurisdictional thresholds should be introduced to evaluate anti-competitive effects of mergers involving Big Data, this thesis compares the merger control thresholds of Germany, the UK, Denmark, the Netherlands and Norway. These MSs were chosen because they each employ a particular jurisdictional test that is able to submit mergers involving Big Data to the merger review process.

Accordingly, this chapter is divided in four sections. Each section gives an overview for each national jurisdictional merger control test that is able to address Big Data concerns. Each section discusses the substance of the test, its legal basis, its motivation, why it addresses Big data concerns and the voluntary or mandatory character of the notification system.

2.1. Germany

In June 2017, Germany introduced a new jurisdictional threshold based on the value of transaction in its existing mandatory notification system. Next to the standing purely turnover thresholds, the GWB’s ninth reform established a notification obligation for concentrations when: a) the combined worldwide turnover of all participating enterprises exceeded EUR 500 million; b) the turnover of one participating undertaking exceeded EUR 25 million within Germany, but neither the target nor any other participating undertaking had a turnover in Germany which exceeded EUR 5 million; c) the value of consideration for the transaction exceeds EUR 400 million; and d) the target is active in Germany to a significant extent (local nexus).12 The German system thus upholds a ‘deal-size’ or ‘transaction-value’ based threshold.

Notification is required when the value of consideration exceeds EUR 400 million, which ensures that only economically important cases are caught.13 The threshold is based on the ‘value of consideration’ for the concentration.14

12_{Section 35(1a) Act against restraints of competition of 26 June 2013 amended on 30 October 2017 (hereafter: ARC/GWB),}

Bundesgesetzblatt I, p. 3618.

13_{W. Berg, ‘Germany proposes transaction value threshold to require notification of high value deals even with no/de}

minimis sales in Germany’, KCLB, 2016.

(9)

Big Data is an intangible asset for companies. The ‘consideration’ encompasses all assets and other monetary benefits that the seller receives in return for the concentration. The term ‘asset’ is interpreted in a broad sense, meaning that it covers all tangible assets (for example, real estate or machinery) and intangible assets (licenses, trademark rights, data).15 Therefore, Big Data is included in the value of consideration for purposes of the transaction value.

The goal of the new threshold was to close the alleged legal gap, i.e. to avoid situations where no notification is needed for acquisitions of companies that are currently achieving only a small turnover, but that can shortly become important in the competitive process.16 This mainly relates to the acquisition of start-up companies, the so-called ‘unicorns’. By impeding undertakings with market power of acquiring prominent start-ups in possession of data that can considerably enhance the market power of that acquirer, innovation and competition in technology markets is protected.17

Germany is not the only state maintaining a ‘size of transaction’ test. The threshold is inspired by the ‘size of transaction’ test used in the US.18_{Austria followed Germany and similarly} launched a ‘size of consideration’ test in combination with lower turnover thresholds to assess the increasing number of high-value deals.19 The Austrian thresholds are not discussed, since they predominantly rely on the German ones.

2.2. United Kingdom (UK)

The UK has its own way of dealing with challenges of the digital world. As opposed to most jurisdictions, the UK keeps a voluntary notification system. A concentration qualifies for review if it meets either of two alternative jurisdictional tests. Besides turnover-based thresholds, the UK retains a ‘share of supply’ test. For purposes of this thesis, only the ‘share of supply’ test is evaluated, since the turnover test does not address Big Data concerns. The ‘share of supply’ test

15_{Bundeskartellamt, Bundeswettbewerbsbehörde, Joint Draft for public consultation, Guidance on Transaction Value}

Thresholds for Mandatory Pre-Merger Notification (Section 35(1a)GWB and Section 9(4)KartG), (2018) (hereafter: BkA Guidance Transaction value), p. 3.

16_{Government Proposal ARC/GWB (see infra fn. 15), p. 1.}

17_{F. Schoening, C. Ritz, ‘Hunting unicorns: German and Austrian Competition Authorities publish draft Guidance Note on}

Transaction Value Thresholds’, Hogan Lovells Focus on Regulation, 2018 (hereafter: F. Schoening, ‘Hunting Unicorns’)

18_{Section 8 Hart-Scott-Rodino Antitrust Improvements Act of 1976 (the HSR Act – USA).}

19_{Section 9 § 4 Austrian Cartel and Competition Act of 2005 as amended on 24 April 2017 (ACA), CELEX-Nr.:}

(10)

is satisfied when the merger itself creates or enhances a 25 per cent share of supply or purchases of any goods or services in the UK, or a substantial part of it.20

The ‘share of supply’ approach is taken because the UK Government wishes to ensure that it can act when a merger raises national security concerns, because it involves a business with a significant share of the supply of critical goods or services in the UK.21 Although this jurisdictional threshold is not specifically meant to address Big Data concerns, this test has enabled the UK Competition Authority (CMA) to investigate competitive effects of several transactions in the digital sector.22

Big Data sets enable companies to improve services or develop new ones. Competition concerns may arise if data aggregation could strengthen the merged entity in a way that increases its share of supply of services or goods. This is especially the case when data is an important asset in the specific market of the goods or services in question. In this hypothesis, the control of Big Data can lead to a sufficient increase in share of supply to meet the notification threshold. Albeit it is debatable whether solely aggregation of Big Data causes the increase in share of supply, it is in this hypothesis that the ‘share of supply’ test can be used to tackle mergers involving vital Big Data. The UK is the only country with a ‘share of supply’ test.

2.3. Denmark and the Netherlands

Another way to address Big Data-related mergers, is by introducing ‘sector-specific’ merger control thresholds, such as Denmark and the Netherlands hold. Both maintain a mandatory merger notification regime.

Denmark works with a ‘sector-specific’ notification threshold for mergers in the telecommunications sector. The second limb of the primary turnover thresholds, i.e. that the turnover of each of at least two of the undertakings concerned should be more than DKK 100 million (approximately EUR 13,4 million) in Denmark,23 has been abolished for the sector of

20_{Section 23A Enterprise Act of 2002 (EA) as amended by the Enterprise and Regulatory Reform Act of 2013 (ERRA) (UK)}

(hereafter: Enterprise Act).

21_{Department for Business, Energy and Industrial Strategy, Draft Guidance on the Enterprise Act 2002: Changes to the}

turnover and share of supply tests for mergers (2018), p. 29.

22_{Public consultation merger control, Reply CMA, § 14.} 23_{Section 12 Danish Competition Act No. 155 of 1 March 2018.}

(11)

public electronic communications networks. If a merger falls within this sector, and the combined aggregate turnover in Denmark of all the undertakings concerned is more than DKK 900 million (roughly EUR 120 million), the threshold is met and the merger must be notified, irrespective of the aggregate turnover of the individual undertakings.24 This causes mergers in this sector to be subject sooner to filing obligations than in other sectors.

The purpose of introducing ‘sector-specific’ merger control was to ensure effective merger legislation in relation to the specific competitive settings existing in the telecommunications market. While the jurisdictional thresholds are still based on turnovers, ‘sector-specific’ merger control regimes allow the focus to remain on sectors and markets where mergers bring about particularly anti-competitive effects.

Where Denmark has special notification requirements for telecom companies, the Netherlands has specific notification thresholds in other specific sectors, such as in the healthcare sector.25 Both maintain a mandatory notification regime.

In a variety of sectors Big Data is an important asset for undertakings and may encompass a notable competitive advantage. Moreover, it is conceivable to introduce ‘sector-specific’ jurisdictional thresholds in sectors and markets where Big Data is a key factor in competition. This holds true in sectors where network effects are present, as they uphold an increasing commercial value by establishing large customer bases. Subsequently, the expansion of the network is achieved by providing products or services at low prices or no price at all. This relates to, for instance, sectors such as high-tech industries, online platforms/advertising, the pharmaceutical sector, etc...

24_{Section 51A amendment to the Danish Electronic Communications Networks and Services Act No. 169 of 3 March 2011}

juncto Section 12 Danish Competition Act No. 155 of 1 March 2018.

25_{Article 29 Act on the new rules on the economic competition of 22 May 1997; Article 49a(1) Act on the rules on market}

organization, efficiency and controlled cost development in the area of healthcare of 7 July 2006; Article 1 Executive order on the temporary extension of the application of merger control on undertakings in the healthcare sector of 1 January 2015.

(12)

Other Member States, viz. Italy, the UK, Ireland and France, also hold ‘sector-specific’ merger control thresholds.26_{However, for purposes of this thesis, the analysis is limited to Denmark} and the Netherlands.

2.4. Norway

Finally, Norway employs a ‘flexible’ merger control system. Norway has possible filing obligations for transactions below the existing mandatory turnover thresholds, subject to a decision by the Norwegian Competition Authority in individual cases or as a more general obligation for individual undertakings.27 Norway employs this regime because it provides a more ‘flexible’ merger control than purely turnover thresholds alone lead to.

The Norwegian CA’s discretionary powers to require notifications below the thresholds, exist mainly as a means of protection of competition in local and niche markets, which could fall outside the scope of the current regime due to higher turnover thresholds.28_{Nonetheless, this} ‘flexible’ merger control system, similarly, allows the CA to address transactions involving high corporate values and Big Data sets.

The Norwegian CA may order a notification for concentrations not meeting the mandatory turnover thresholds, if there are ‘reasonable grounds to assume that competition will be affected’ by the concentration. This order is subject to a strict deadline of three months after the final agreement is concluded or control is obtained, whichever comes first.29 Hence, the CA must keep an eye on certain transactions that involve Big Data and can reasonably affect competition. Subsequently the Norwegian CA can order notification of those transactions within strict time limits and further investigate the anti-competitiveness of the deals.

26_{J.-F. Bellis, P. Elliott (editors), Merger control: jurisdictional comparisons (2}nd_{edition, European Lawyer Reference),}

2014 (hereafter: J.-F. Bellis, Merger control).

27_{Sections 16 to 21 of the Norwegian Competition Act of 5 March 2004.} 28_{T.S. Skaug, ‘Norway’ in J.-F. Bellis, Merger control, supra fn. 26, p. 589-590.} 29_{Section 18(3) Norwegian Competition Act.}

(13)

3. How do the five national merger control systems comply with the normative benchmarks?

The five jurisdictions are compared according to four normative benchmarks. These benchmarks create essential guiding principles in assessing the added value of additional jurisdictional thresholds for the EU merger control system. Based on, the EUMR, the ICN Recommended Practices and the doctrine, the following four benchmarks are fundamental to shape a proper merger control system tackling Big Data concerns: 1) degree of tackling Big Data concerns 2) legal certainty, 3) cost-effectiveness of the merger control system, and 4) a sufficient local nexus.

Additionally, it should be reminded that the EU, consisting out of 28 (or soon to be 27) Member States, is not comparable to solely one Member State. This means that difficulties may arise if national merger control thresholds are applied in the EU framework. Consequently, the analysis of the various benchmarks, also considers the peculiarities of the EU system. In this regard, new jurisdictional thresholds should respect the principles of subsidiarity30 and proportionality31, which remain of substantial importance in the EU merger control system in order to safeguard the ‘one-stop-shop’ system.32

This chapter exists out of five sections. Four sections deal with the four benchmarks. Each section starts with a brief introduction about the benchmark in question and why it is important for this thesis. Furthermore, the five chosen jurisdictions are evaluated according to the particular benchmark. Finally, each section ends with a conclusion, which is based on a ranking of the jurisdictions according to the evaluated benchmark. The final rankings are correspondingly included in Annex I. The final section gives a synthesis of the analysis of each jurisdiction based on the four benchmarks.

30_{Article 5(3) TFEU.}

31_{Article 5(4) TFEU; Protocol No. 2 to the TFEU of 09 May 2005 on the application of the principles of subsidiarity and}

proportionality.

(14)

3.1. Degree of tackling Big Data concerns

Chapter 2 assessed the different ways the chosen national merger control systems are able to tackle mergers raising Big Data concerns. However, the degree to which Big Data concerns are dealt with, may vary from one kind of threshold to another. Consequently, not all five jurisdictions are able to achieve the objective of tackling Big Data concerns equally effective.

In this regard, this section assesses the degree to which the five jurisdictions tackle Big Data concerns, on the basis of various subtests. The first test is whether Big Data is taken into account directly or indirectly (the more direct, the higher the degree of tackling the Big Data concerns). Secondly, it is analyzed whether the Big Data concerns are fully or partially tackled (the more fully, the higher degree of tackling Big Data concerns). Finally, the question is raised whether the Competition Authority is left with discretionary powers to decide upon the submission of mergers involving Big Data to merger review process (the more discretionary powers, the lower the degree of tackling Big Data concerns).

3.1.1. Germany

The ‘size of transaction’ test is able to take Big Data in account, since it forms an intangible asset for companies, which is taken into account for the value of consideration. Consequently, Big Data is taken directly and fully into account in the merger notification threshold. Furthermore, Big Data concerns is taken into account, even without the use of discretionary powers on part of the CA, since it is up to the undertakings concerned to determine the value of the transaction, including the Big Data involved, and, subsequently their notification requirement.

3.1.2. United Kingdom

A voluntary notification system combined with a ‘share of supply’ test could take into account Big Data issues in the hypothesis that the data strengthens the market power of the entity in such a way that it increases its share of supply in the specified goods or services. This will depend on several factors concerning the market involved. Consequently, the ‘share of supply’ test only takes Big Data indirectly into account. Additionally, Big Data concerns are not taken fully into account, since the needed increase of supply may not be exceeded, even where Big

(15)

Data is involved. The ‘share of supply’ test is consequently not an adequate solution to tackle Big Data concerns in all markets where Big Data presumably creates a competition issue. Nevertheless, Big Data is taken into account, even without the use of discretionary powers on part of the CA, since the ‘share of supply’ test is objectively calculated.

3.1.3. Denmark and the Netherlands

Denmark and the Netherlands have lowered their turnover thresholds in specific sectors. Hence it is possible to lower existing turnover thresholds to tackle Big Data issues in sectors where Big Data is a key factor in competition and, therefore, particularly can create anti-competitive concerns. This relates to, for instance, sectors such as high-tech industries, online platforms/advertising, the pharmaceutical sector, etc... However, it might be hard to identify each sector throughout the EU where data is sufficiently important to justify the need for ‘sector-specific’ thresholds.

Consequently, Big Data is only taken indirectly into account, i.e. by determining the sectors in which they particularly may raise anti-competitive concerns. Further, Big Data concerns are tackled only partially, since Big Data is mostly not reflected in the turnovers of the companies involved.33 Accordingly, in order to tackle Big Data concerns, the turnover thresholds need to be sufficiently low. Finally, since turnover is objectively calculated, there is no need to exercise a discretionary power on part of the CA.

3.1.4. Norway

The ‘flexible’ merger notification system, gives the NCA the opportunity to tackle every merger below the turnover thresholds, that it considers to be raising competition concerns. This enables the CA to identify mergers involving Big Data. However, it requires the CA to constantly monitor the market to identify potentially harmful mergers.

Such a system may take Big Data directly and fully into account, since the CA can focus specifically on mergers involving Big Data. However, the CA employs substantial discretionary powers to decide whether mergers involving Big Data are submitted to the merger review

(16)

process. In this regard, it will depend on the CA whether the focus will lie on Big Data concerns and whether these will be addressed.

3.1.5. Conclusion on the degree of tackling Big Data concerns

Jurisdiction Degree of tackling Big Data concerns

Germany: ‘size of transaction’ - Direct - Full - No discretionary powers The UK: ‘share of supply’ - Indirect - Partial - No discretionary powers

Denmark & the Netherlands: ‘Sector-specific thresholds’

- Indirect - Partial

- No discretionary powers

Norway:

‘Flexible’ merger control system’

- Direct - Full

- Discretionary powers

Ranking

( 1 being the highest ranked)

1. Size of transaction 2. Flexible system

3. Share of supply

4. Sector-specific thresholds (Table 1)

It can be concluded that the German ‘size of transaction’ test maintains the highest ability to tackle Big Data concerns, since it directly and fully takes into account Big Data, without the need for intervention by the CA. Opposed to the German system, the ‘flexible’ Norwegian system necessitates discretionary powers to consider Big Data concerns fully and directly. Conversely, both ‘sector-specific’ thresholds as the ‘share of supply’ test do not require discretionary powers upon the CA to deal with Big Data concerns. However, they both only cope indirectly and partially with Big Data issues. In this regard, the ‘share of supply’ test ensures a slightly higher degree of tackling Big Data concerns, since the possession of Big Data may lay at the root of the increase in supply share, while it is harder to specifically connect turnover in the ‘sector-specific’ thresholds to the possession of Big Data.

(17)

3.2. Legal certainty

Legal certainty is a fundamental principle, recognized by most jurisdictions across the world. The principle contributes to the rule of law and constitutes a requirement for efficiency in market interactions.34 In merger control systems, legal certainty aims, inter alia, to preserve the legitimate expectations of undertakings contemplating a concentration.35 Commissioner Vestager confirmed that it is a fundamental in EU merger control that there is no doubt regarding whether a merger must be notified.36 For the effectiveness, the subsidiarity and the proportionality of the EUMR, it is crucial that legal certainty exists regarding which transactions are subject to the notification requirement and the standstill obligation. Likewise, legal certainty is referred to in the EUMR several times.37 It is also perceived as essential by the ICN Recommended Practices. Consequently, new EU merger control thresholds should guarantee a high level of legal certainty.

To assess whether jurisdictional thresholds fulfill the highly valued EU requirement of legal certainty, different aspects play a role. In brief, these aspects are: 1) the clarity of the concepts used in the thresholds, 2) whether or not they are based on ‘objectively quantifiable criteria’, 3) whether the jurisdiction maintains a mandatory or voluntary notification system, and finally 4) whether the thresholds are accompanied with adequate guidelines. Hereafter follows a short explanation on these aspects.

First, according to the ICN Recommended Practices, notification thresholds should be ‘clear, understandable, easily administrable, bright-line’ tests, based on ‘objectively quantifiable’ criteria.38

To meet the standard of ‘objectively quantifiable’ criteria, criteria must refer to assets and sales or turnover. Unlike turnover, market share and potential transaction related effects, are not considered to be objectively quantifiable. Market share based tests and transaction related tests

34_{A. Portuese. ‘The principle of legal certainty as a principle of economic efficiency’, European Journal of Law and}

Economics, 2014, p. 1.

35_{UNCTAD ‘Enhancing legal certainty in the relationship between Competition Authorities and Judiciaries’,}

TD/B/C.I/CLP/37 (2016).

36_{Speech Commissioner Vestager ‘Refining the EU merger control system’ (2016).} 37_{Recitals 11, 25 and 34 EU Merger Regulation.}

(18)

are facts-based and inherently subjective, thus fundamentally conflicting with the need for undertakings to ascertain legal requirements in time-constrained circumstances.39

The next aspect playing a role in the assessment of legal certainty, is whether the merger control is within a mandatory notification system (possibly supplemented with additional notification thresholds) or a voluntary notification system.

In a system with mandatory notification thresholds, the parties to the merger enjoy at least a degree of legal certainty. When the mandatory notification thresholds are exceed, there is complete certainty about the notification obligations. Germany, Denmark and the Netherlands hold mandatory notification systems and, therefore, already guarantee a minimum degree of legal certainty.

Additional to mandatory notification thresholds, jurisdictions may retain the ability to review transactions that do not meet the mandatory notification thresholds. This holds true in Norway, where the Norwegian Competition Authority retains the possibility to review transactions below the mandatory notification thresholds.

Generally, such ‘residual jurisdiction’ may incorporate all transactions with a material nexus to the jurisdiction that meet lower, non-mandatory notification thresholds. However, Norway does not employ such lower non-mandatory thresholds. Residual jurisdiction can diminish the degree of legal certainty. Therefore, it is necessary to include adequate safeguards to maintain a sufficient level of legal certainty, for example by restricting the CA’s ability to exercise residual jurisdiction to a specified, limited period after the conclusion of a transaction. Additionally, the CA could allow the parties to submit voluntary notifications.40_{Finally, jurisdictions can} enhance legal certainty by publishing guidelines on how the additional notification thresholds will be applied.

By contrast, jurisdictions could not employ mandatory notification thresholds and, instead, choose for voluntary notifications of proposed transactions. Voluntary jurisdictions can thereby

39_{Ibid, Section IIE, Comment No. 1, p. 6.} 40_{Ibid, Section IIA, Comment No. 3, p. 3.}

(19)

decide to employ certain thresholds. On one hand, these thresholds provide guidance to merging when concentrations should be notified, due to the likelihood of raising potential competition concerns. This is the case in the United Kingdom, where the ‘share of supply’ test is one of the possible thresholds that must be exceeded to qualify for review. On the other hand, voluntary thresholds can limit the transactions the CA can review. Voluntary merger notification systems should, like ‘residual jurisdiction’, embrace adequate safeguards to address the wish of the undertakings concerned for certainty.41

Furthermore, voluntary or residual (supra) merger regimes may infer extra difficulties with respect to the decision time upon the approval of the merger, the need for informal advice or preliminary discussions and the possibility of imposing interim measures. Consequently, these aspects will also be assessed in case the evaluated jurisdiction maintains a voluntary or residual merger notification system.

3.2.1. Germany

Germany upholds a mandatory notification system, which already embraces some degree of legal certainty. However, at least two of the requirements of the German jurisdictional thresholds raise difficulties regarding legal certainty. These are that a) the amount of ‘the consideration for the transaction’ must exceed EUR 400 million, and that b) the undertaking to be acquired is ‘significantly active’ in Germany. The difficulties regarding legal certainty are caused by the vagueness of the concepts used in these thresholds, and because the ‘consideration for the transaction’ can hardly be observed as an ‘objectively quantifiable’ criterium.

Although ‘consideration of the transaction’ is defined to encompass all assets and other monetary values received by the seller in return for the concentration, the concept remains ambiguous and inherently raises various complications regarding legal certainty. It is the merging parties’ responsibility to check the value of consideration and determine their

(20)

notification obligayion.42_{However, merging parties face multiple complexities in determining} both the overall value of consideration, as well as, the value of the Big Data involved.

First, parties might still have to decide on the value of the data and the overall transaction when the intention of merging is notified. Similarly, parties could have made use of complex price formulas or constructions with options. This implies that the true value is not immediately apparent upfront.43 Consequently, determining the notification requirement, is not an easy assignment.44

Secondly, defining a monetary value for data is a challenging task. This is due to various reasons such as difficulty in delineating the value of data from the processes that generate and treat data. This is the case especially if the undertaking collecting the data is not yet fully monetizing the data.45

Moreover, value is inherently subjective, and can change materially over short periods, as it is subject to opinion and market volatility.46 Therefore, it is burdensome to quantify the value of the data and the transaction in general. Facing difficulties in determining the value of data and the concentration can, consequently, raise doubts concerning the filing obligations. This evidently fails to ensure legal certainty.

Furthermore, although initially it is up to the undertakings concerned to determine the value of the transaction and subsequently the notification requirement, the BkA can take a different view regarding the value of the data and the overall transaction. Consequently, the BkA may call for a notification, where the merging parties did not envisage it or concluded that there was no obligation. In the latter situation, they risk being sued for infringing the stand-still obligation. Hence, the parties are recommended to notify as a precaution in cases of doubt.47_{. This shows}

42_{BkA Guidance Transaction value p. 6.}

43_{It can also not be ruled out that the undertakings concerned employ other complex payment structures or set artificial}

values to disguise the real economic value of the transaction; Public consultation merger control, Reply Antimonopoly Office of the Slovak Republic; Public consultation merger control, Reply Catalan Competition Authority.

44_{F. Schoening, ‘Hunting Unicorns’, supra fn. 17.}

45_{A. Boutin, X. Boutin, ‘Opinion paper on merger control: Proposals for a more efficient European Merger Control’,}

ECARES, 2017, p. 14.

46_{I. Kokkoris, Merger control in Europe: the gap in the ECMR and national merger legislations (Routledge), 2011, p. 69-70;}

Public consultation merger control, Reply Simmons & Simmons LLP.

(21)

that the ‘size of transaction’ test causes uncertainties for the parties and, therefore, fails to meet the legal certainty requirements.

Finally, the value of consideration is calculated on the date of completion of the merger. However, the value of consideration and of the data, can fluctuate over time. This causes a problem when the value increases to an extent that the thresholds are exceeded at the merger completion date, while they were not at time of assessment of the notification obligation. In this case, an obligation to notify arises before the completion date. This evidently creates legal uncertainty the period between assessment and completion date. Accordingly, the merging parties are advised to notify their transaction as a precaution. Subsequently, the ultimate responsibility for reviewing the notification requirement lies with the CA.48 If the CA eventually decides that no notification was needed, the parties can withdraw their application.

The last limb of the German thresholds, similarly raises concerns regarding legal certainty. It requires that one of the undertakings, that does not achieve EUR 25 million turnover in Germany, i.e. in most cases the target company, is ‘significantly active’ in Germany. This qualifies as the local nexus requirement (infra). The law as such remains silent on the criteria to determine when an undertaking is ‘significantly active’ or ‘active in a large scale’ in Germany. The Government proposal and the BkA Guidance on transaction value clarify the meaning of ‘significantly active’ to some extent (infra). However, these guidelines remain to encompass ambiguous concepts and apply different criteria to different sectors and activities, both concerning ‘domestic activity’ as ‘significance’. In absence of clear criteria regarding the local nexus, the thresholds unfortunately generate an element of legal uncertainty in the jurisdictional assessment.49

The UK merger control system raises predominantly two issues regarding legal certainty. First, because it holds a voluntary merger control regime. Secondly, because the ‘share of supply’ is hardly an objectively quantifiable criterium as advised by the ICN Recommended Practices.

48_{Ibid, p 7.}

49_{Cleary Gottlieb Alert Memorandum, ‘Germany and Austria introduce Transaction Value Merger Notification Thresholds’}

(22)

Notwithstanding, the UK does try to compensate the issues around legal certainty with guidelines and the possibility of informal discussions.

The UK retains a voluntary merger control system, in the sense that there is no obligation to apply for CMA clearance before completing a transaction. The question whether clearance should be pursued in a particular case is for the parties to consider, which comes down to an assessment of the commercial risk of not notifying.50 Legal certainty in a voluntary notification system is hard to accomplish. Therefore, since legal certainty is highly valuated in EU law, such a system would not stand throughout the Union.

According to the ICN, notification thresholds in voluntary notification systems, serve to provide guidance on which transactions are observed likely to raise competition concerns. Therefore, it is appropriate for voluntary notification thresholds to employ objective criteria and provide guidance to assist parties in determining which transactions meet the thresholds.51

Furthermore, the concept of ‘share of supply’ is complicated and may necessitate an appraisal on behalf of the CMA. The merging parties cannot easily identify when the test is met, as they lack sufficient market information.52 This does not benefit legal certainty.

Nevertheless, merging parties can ask for informal advice or engage in preliminary discussions with the CMA about the application of the ‘share of supply’ test, to increase their legal certainty. Hereby the CMA is willing to give informal indications that it will not review a merger.53 The CMA and OFT have also issued guidelines on how to interpret the CMA’s jurisdiction and the ‘share of supply’ test.54

The hypothesis could be made to introduce the ‘share of supply’ test in the mandatory EU notification system. Albeit pre-notification discussions, informal advice and extensive guidelines on the jurisdictional thresholds, assessing whether the ‘share of supply’ test is met, is complicated and requires good expertise. This often requires a degree of judgment on part of

50_{Slaughter and May, UK Merger control under the Enterprise Act 2002 (2016).} 51_{ICN RP, Section IIE, Comment No. 4, p. 6.}

52_{Public consultation merger control, Reply CMA (UK).}

53_{Paragraph 6.22 et seq. CMA (UK) Mergers: Guidance on the CMA’s jurisdiction and procedure (2014).} 54_{Ibid, Paragraph 4.53 et seq.; OFT and Competition Commission, Merger assessment guidelines (2010).}

(23)

the European Commission. This is more challenging and burdensome within a mandatory merger control system, where even more certainty is required about whether the thresholds are exceeded, since the merging parties risk being sanctioned for ‘gun-jumping’.55_Additional all-encompassing guidelines and preliminary help of the CA in calculating the ‘share of supply’, could contribute in resolving these issues.

Furthermore, when a merger meets the ‘share of supply’ test, it qualifies for review by the CMA. If the CMA deems the transaction to result or expects to result in a substantial lessening of competition in a UK market, then it will refer the merger for Phase-II Investigations. The Phase-II reference is no longer possible four months after implementation.56 This leaves the merging parties in uncertainty during that period.

Additionally, the four months do not start until the ‘material facts’ of the merger are made public or given to the CMA. Time also not runs when the parties are negotiating, are yet to comply with an information request from the CMA, or when the CMA made a request for review of the transaction in accordance with Article 22(3) EUMR.57 All these situations prolong the uncertainty period for the undertakings involved, who await whether the transaction is allowed to remain.

As mentioned before, the CMA can impose interim orders during the substantive assessment. It can stop any integration that might constitute pre-emptive action. It may also require the parties to undo any integration steps that were already done. In conclusion, although the UK overall tries to ensure legal certainty with informal advice, preliminary discussions and guidelines, legal certainty is far from accomplished.

It is questionable whether ‘sector-specific’ merger notification thresholds contribute to legal certainty or fairly the opposite. Both Denmark and the Netherlands lowered their turnover thresholds for specific sectors. According to the ICN, turnover thresholds are considered

55_Ibid.

56_{Section 24 Enterprise Act.}

(24)

generally clear, understandable, easily and objectively quantifiable.58_{Merging parties can} easily assess whether their turnovers meet the thresholds. Placed in a mandatory notification regime, which is the case in Denmark and the Netherlands, ‘sector-specific’ thresholds preserve already at least a minimum degree of legal certainty.

Nevertheless, the fact that different thresholds apply to different sectors, can lead to several uncertainties. First, the definition of the sectors they apply to, must be concise. In fact, it can be problematic to identify the boundaries between specific industries with sufficient clarity. Therefore, it is recommended to have adequate guidelines elucidating the delimitation of the sectors, so that undertakings are fully capable to assess their notification obligations. In this respect, both Denmark and the Netherlands are doing well, by giving clear definitions of electronic communications (providers)59 and healthcare (providers) respectively.60

3.2.4. Norway

The ‘flexible’ Norwegian merger control system hardly realizes legal certainty, because of mainly three reasons. First, Norway works with a mandatory notification system accompanied with residual jurisdiction (supra). Secondly, it is unclear when a merger may be subjected to a notification order. Thirdly, Norway fails to provide sufficient guidelines to compensate the former problems.

Contrary to most ‘residual jurisdiction’ systems, Norway has no sign of any thresholds indicating when a concentration would be subjected to a notification order by the Norwegian CA. When the transaction does not fulfill the general mandatory turnover thresholds, the proposed concentration may proceed without being subject to a stand-still obligation. Consequently, merging parties endure three months of uncertainty about their notification obligations. In order to attain faster illumination, parties may choose to engage in preliminary discussions or submit a notification as a precaution.61

58_{ICN RP, Section IIE, p. 6.}

59_{Section 2 Danish Electronic Communications Networks and Services Act No. 169 of 3 March 2011.}

60_{Dutch Act on the rules on market organization, efficiency and controlled cost development in the area of healthcare of 7}

July 2006.

(25)

To secure legal certainty, the NCA may only intervene in transactions where it is reasonable to assume competition is affected or where other special considerations require further investigation. If the Authority chooses to intervene, it must give a reasoned statement as to why these criteria are fulfilled in the transaction. Additionally, the option to intervene below the notification thresholds is emphasized as a restricted one.62 Albeit these considerations, the risk of subjectivity on part of the CA is plausible.

In addition, when the Norwegian CA decides to order notification, parties may have implemented the concentration. After all, the CA can mandate notification up to three months after the agreement is concluded or control is obtained, whichever comes first. Although the limited time frame tries to protect legal certainty, the exact period remains unclear, since ‘conclusion of an agreement’ is an ambiguous term in the merger framework. Also, the fact that the three months can start on two distinct occasions ‘whichever comes first’ can cause uncertainties.

Furthermore, during the CA’s investigations, normal procedural rules apply. Hence, considering a completed merger, the undertakings involved must freeze any steps in implementing the transaction. Moreover, the CA can determine the transaction anti-competitive. Consequently, it has the power to reverse all measures carried out by the parties, and/or subject the transaction to conditions. This raises the level of uncertainty significantly. Therefore, it is important for the partiesn in a proposed concentration falling below the mandatory thresholds, to assess the risk of being caught by the CA, and consider a precautionary notification.63

3.2.5. Conclusion on legal certainty

Jurisdiction Legal certainty

Germany:

‘size of transaction’

- Mandatory notification system - Clarity of the concepts:

 Value of consideration  Ambiguous concept

 Significantly active in Germany  Ambiguous concept

 Different criteria for different industries - Objectively quantifiable criteria:

62_Ibid. 63_Ibid.

(26)

 Subjectivity of value (different view by the CA)  Defining monetary value for Data

 Time of determination of the value is crucial - Guidelines

 Yes, but still ambiguous

The UK: ‘share of supply’

- Voluntary notification system  Issues:

 Less legal certainty compared to mandatory systems  Demand a notification up to 4 months after implementation  Possibility to prolong the limitation period

 Need for informal advice or preliminary discussions  Risks of interim and reverse measures

- Clarity of the concepts  ‘share of supply’  Complex

- Objectively quantifiable criteria:  ‘share of supply’ hard to quantify  Complex to calculate

 Merging parties lack market information to calculate - Guidelines

 Yes, extensive

Denmark & the Netherlands: ‘Sector-specific thresholds’

- Mandatory notification system - Clarity of the concepts

 ‘Sector-specific’ turnover thresholds  Turnover thresholds contribute to certainty  Need for clear criteria delimiting the sectors - Objectively quantifiable criteria:

 ‘Sector-specific’ turnover thresholds: yes - Guidelines

 Yes, extensive

Norway:

‘Flexible’ merger control system

- Residual notification system  Issues

 Less legal certainty than fully mandatory system  Demand of notification up to 3 months after conclusion

agreement

 Need for informal advice or preliminary discussions  Risks of interim and reverse measures

- Clarity of concepts

 No indication of thresholds - Objectively quantifiable criteria:

 ‘Reasonable to assume competition is affected’  Risk of subjectivity by the CA

- Guidelines  Insufficient

Ranking

( 1 being the highest ranked)

1. Sector-specific thresholds 2. Size of transaction 3. Share of supply

4. Flexible system (Table 2)

New EU merger control thresholds that address Big Data concerns should adhere to the highly valued principle of legal certainty. In comparison of the four jurisdictional tests, legal certainty is best preserved in a system with ‘sector-specific’ merger control thresholds, since turnover is

(27)

used as the standard. This is the case in both Denmark as the Netherlands. Turnover is clear and objectively quantifiable, while the ‘value of consideration’ in the German jurisdictional thresholds is predominantly ambiguous, rather subjective and essentially difficult to determine. Also the ‘share of supply’ test is, opposed to turnover, hard to verify, since merging parties lack substantial information regarding the market. Additionally, the voluntary notification system handled in the UK brings about less legal certainty than mandatory notification systems, as there is no specific obligation to notify. Moreover, a voluntary notification system, like a ‘flexible’ merger control system, involves the risk of being subjected to interim measures and reverse measures, which could be detrimental for already implemented concentrations. Finally, merging parties in the Norwegian ‘flexible’ system retain least legal certainty, since there are no indications of when to notify. Accordingly, the CA is in the driving seat in such a system, since the conditions to order notifications are ambiguous and can lead to subjectivity on the CA’s side.

(28)

3.3. Cost-effectiveness

Merger notifications impose abundant transaction costs, and demands substantial commitment of a CA’s resources. The process of notification also levies abundant costs on the merging parties. To evade redundant expenditure of public and private resources, the merger notification thresholds must screen out transactions that are unlikely to result in appreciable competitive effects within a jurisdiction.64 ²

It is commonly recognized that setting and adjusting notification thresholds should balance between the desire to review most transactions that may harm competition, and the need to keep the process and costs manageable, predictable and reasonable for all sides involved.65 In order to have a cost-effective system, jurisdictional thresholds should be adapted to the CAs’ capacity. The merger control thresholds should be focused to contribute to an efficient, timely and effective merger review, without imposing unnecessary transaction costs or loss of resources.66

Jurisdictions should pursue to ensure, without limiting the effectiveness of merger control, that notification requirements do not impose unnecessary costs and burdens on merging parties and third parties.67 Hence jurisdictional thresholds should be clear and objective, which also refers to legal certainty. They should avoid notifications of nonthreatening mergers.68

New EU merger control thresholds must safeguard a cost-effective EU merger control system. Hence, this section analyzes whether the five jurisdictions hold a cost-effective merger control system, considering in particular; 1) the capacity of CAs to handle the merger control process, 2) the need for extra capacity to monitor the market, 3) time and resources spent on preliminary discussions and 4) costs and burdens merger proceedings may incur on merging parties.

64_{ICN RP, Section IIB, p. 3.}

65_{ICN Setting Notification Thresholds for Merger Review (2008), p. 4; OECD Secretariat Background paper ‘Local Nexus}

and jurisdictional thresholds in merger control’ (2016), p. 5.

66_{ICN NP Working Group, Guiding Principles for Merger Notification and Review (2002).} 67_{OECD Recommendation C(2005)34 of the Council on Merger Review (2005), p. 2.} 68_{Ibid, p. 3.}

(29)

3.3.1. Germany

Since the new German thresholds only came into force last year, there is not yet an official publication of statistics on their application. However, the BkA did confirm that they only received a low number of notifications on account of the new thresholds. In most instances, the thresholds were not met and were subsequently followed by a withdrawal of the notification. In other cases, the question whether the thresholds were met, was left open, since the cases did not pose any substantial issues. None of the cases went into Phase II of the merger proceedings. Due to this merely small increase of cases, the workload of the BkA did not considerably increase. Many transactions captured by the new thresholds, are likely caught already by the existing turnover thresholds. 69

However, one should not lose sight of informal preliminary discussions before it comes to actual cases. The additional thresholds create an increased need for merging parties to coordinate with the BkA, since they not want to risk failing a potential notification obligation in Germany. As there are still numerous uncertainties about the interpretation of the new thresholds, pre-notification discussions will occur rather frequently. With this background, it is questionable whether the BkA has sufficient capacity to handle the combination of preliminary discussions with an actual increase of new merger cases.

That leaves the question whether the additional ‘size of transaction’ threshold causes unnecessary costs and burdens on merging parties and third parties. As mentioned above, the ‘size of transaction’ test comprehends some vague legal concepts such as ‘consideration’ and ‘significant active’. This ambiguity can lead to increased costs and burdens for merging parties, because they face difficulties in deciding their filing obligations. To be certain, merging parties could engage in preliminary discussions with the BkA. However, this is both time-consuming and costly. According to the BkA, already some precautionary notifications occurred which in fact were not compulsory. The probability that this will occur even more in the future is by and large conceivable. To prevent these additional costs and burdens, it is recommended to elucidate on ambiguous concepts and provide sufficient guidelines.

69_{F. Schoening, C. Wünschmann, C. Ritz, ‘Mind the gap - New size-of-transaction test in German merger control’, Hogan}

(30)

Since the UK merger control system relies upon voluntary notifications, it would be expected that less mergers are notified and, therefore, the CMA partakes less workload compared to mandatory notification systems. However, the need for capacity may not be underestimated. Such a system demands abundant resources for monitoring, collecting and analyzing information on market developments to identify potentially problematic mergers.

As mentioned before, a voluntary system also triggers preliminary discussions or requests for informal advice by the merging parties, who want to ascertain their obligation to notify. Though, to make useful use of resources, the CMA has stated that it only gives informal advice when it is useful and appropriate, i.e. when there is a genuine issue. It will not give advice in cases that do not evidently raise competition concerns.70 Moreover, as preliminary discussions demand time and resources, the CMA only engages in discussions when the parties wish to proceed to notify the merger.71

Applying the ‘share of supply’ test in the voluntary system can also bring additional costs for the merging parties. To assess the filing obligation, merging parties need to calculate whether the transaction will create or enhance a 25 per cent share of supply of goods, regardless whether the transaction ultimately requires notification. However, parties usually do not own data on the supply market and may lack the ability to calculate accurately their share of supply of the specified goods. Accordingly, this complicated assessment may necessitate involvement of experts, which generates extra costs. Costs also arise as a result of examination and collection of data, and restriction and diversion of business during and because of ongoing procedures. It eventually can cause time delays in completion of the transaction.72

Another option is to ask the CMA for informal advice or engage in preliminary discussions. Both cost extra time and resources on behalf of the companies. There is also an added risk that parties notify transactions purely by way of precaution. This results in a waste of resources when the thresholds were ultimately not met. Notwithstanding, when the merging parties decide

70_{Paragraph 6.28 and 6.33 CMA (UK) Mergers: Guidance on the CMA’s jurisdiction and procedure (2014).} 71_{Ibid, Paragraph 6.24 and 6.39.}

(31)

not to notify, they risk being caught by the CMA eventually. This goes together with the burdens and costs of interim measures, stand-still obligations and even reverse measures.

The inclusion of ‘sector-specific’ merger control thresholds in a merger control system evidently creates extra workload for a CA. By decreasing the turnover thresholds in certain sectors, it is presumed that the number of merger notifications within the specified sectors increases as opposed to merger control with solely universal notification thresholds. Especially in case of an accumulation of various thresholds for multiple sectors, the capacity of the CA can become under pressure.

Therefore, it is questionable whether the European Commission holds sufficient capacity to introduce extra EU ‘sector-specific’ thresholds. Moreover, it should be noted that the European Commission already has extensive sector inquiry powers, which makes it possible to examine and keep in check anticompetitive outcomes from transactions in specific sectors.73

Furthermore, ‘sector-specific’ merger control thresholds increase the notification obligations for companies in specific industries. This necessitates extra vigilance on part of undertakings in these industries regarding their filing obligations, since they are sooner subject to filing obligations than companies from other sectors.

3.3.4. Norway

While in general the task of assessing notification obligations rests upon the undertakings concerned, the Norwegian ‘flexible’ system allows the CA itself to order notifications of concentrations not meeting the turnover thresholds. Hence, it is the task of the CA to keep track of transactions involving Big Data. This additional task of being a ‘watchdog’, creates an extra burden on the CA and might weaken its capacity.

73_{Article 17 Council Regulation (EC) No. 1/2003 of 16 December 2002 on the implementation of the rules on competition}