EUR-Lex Access to European Union law

Back to EUR-Lex homepage

This document is an excerpt from the EUR-Lex website

Document 52000DC0399

Report from the Commission to the Council on the application of the Merger Regulation thresholds

/* COM/2000/0399 final */

52000DC0399

Report from the Commission to the Council on the application of the Merger Regulation thresholds /* COM/2000/0399 final */


REPORT FROM THE COMMISSION TO THE COUNCIL on the application of the Merger Regulation thresholds

REPORT FROM THE COMMISSION TO THE COUNCIL

on the application of the Merger Regulation thresholds

BACKGROUND

1. Council Regulation (EEC) No 4064/89, "the Merger Regulation" was adopted in December 1989 and entered into force in September 1990. One of the main principles of the Merger Regulation was that large concentrations, the market impact of which could be presumed to go beyond any single national market, should be subject to the Commission's exclusive jurisdiction. The criteria for defining concentrations with a Community dimension were based on quantitative turnover thresholds, set out in Article 1 of the Regulation.

2. The concept that the Commission should have sole competence to deal with Community dimension mergers also follows from the principle of subsidiarity. From the viewpoint of the European business community, the Commission's exclusive jurisdiction also provided a one-stop shop principle, which was widely regarded as an essential part of keeping the regulatory costs associated with cross-border transactions at a reasonable level.

3. In addition, the Commission's exclusive jurisdiction to vet such mergers was seen as an important element in providing a level playing field for the concentrations that were bound to result from the completion of the internal market. This principle was widely accepted as the most efficient way of ensuring that all mergers with a significant cross-border impact would be subject to a uniform set of rules.

4. Already from the outset, the turnover thresholds of Article 1 of the Merger Regulation were intended to be reviewed in 1993. At that stage, however, it was considered that insufficient experience had been gained. The review of the thresholds was therefore postponed. In 1998, after a careful review of the experience gained, the Merger Regulation was amended, through Council Regulation 1310/97. The amendments concerned a large number of areas of the Regulation. In relation to Article 1, a new sub-paragraph (Article 1(3)), was introduced. Its intention was to provide a solution to a problem which had been discovered in the course of the review, namely that a significant number of cases that failed to meet the turnover requirements of Article 1(2) had to be notified in a number of Member States ("multiple filings"). This evidently meant that a number of concentrations that did not meet the relatively high threshold in Article 1(2), but which still had a significant cross-border impact did not benefit from the one-stop shop principle. Consequently, there were indications that the Merger Regulation had not fully succeeded in creating the level playing field that it had sought to attain.

5. It may be argued that the most apparent means to overcome the difficulty of multiple filings would have been to provide for Commission jurisdiction in cases where otherwise a notification would have been required to two or more Member States. This apparently simple solution was, however, not available. The reasons for this was that the national rules on merger control still diverged to a large extent. Among the more important limitations to the indicated solution was the fact that some Member States did not have any merger control system and that others had a system based on voluntary notifications. In these circumstances it was considered that relying on national notification requirements as such to establish Community jurisdiction would have been overly complex and risked creating ambiguities as to where specific transactions should be notified.

6. For these reasons, the solution introduced by Council Regulation 1310/97 was to introduce a new set of lower turnover thresholds, intended to achieve the same result (Article 1(3) is set out below). Article 1(4) of the Merger Regulation requires the Commission to report to the Council before 1 July 2000 on the operation of the thresholds and criteria set out in paragraph 1(2) (the original thresholds) and 1(3) (the additional, lower thresholds).

7. The analysis of the operation of the existing turnover thresholds is necessarily of technical nature. Therefore the main purpose of this report can be characterised as an analysis of whether the existing turnover thresholds are appropriate to determine which concentrations have a Community dimension. In addition to the technical analysis, it may be appropriate to carry out a more thorough inventory of the Community merger control system, in order to assess whether or not the existing system is well equipped to face the challenges of the foreseeable future. These challenges will include external factors, such as the extension of the Community through the accession of the applicant countries and the continuing "merger boom", as well as internal factors, such as the modernisation of Community anti-trust rules [1].

[1] See points 79-81 of the Commission's White paper on modernisation of the rules implementing Articles 85 and 86 [now 81 and 82] of the EC Treaty - Commission programme No. 99/027 - published on 28.04.1999.

The existing thresholds

8. As indicated above, the turnover thresholds of the Merger Regulation serve to divide the competence to assess mergers between the Commission and the National Competition Authorities ("NCA's"). Article 1(2) reads as follows:

For the purposes of this Regulation, a concentration has a Community dimension where:

(a) the combined aggregate worldwide turnover of all the undertakings concerned is more than ECU 5 000 million; and

(b) the aggregate Community-wide turnover of each of at least two of the undertakings concerned is more than ECU 250 million, unless each of the undertakings concerned achieves more than two-thirds of its aggregate Community-wide turnover within one and the same Member State.

9. Article 1(3), which entered into force in 1998, extends the Commission's sole competence to assess certain mergers that involve lower turnovers. As indicated above, Article 1(3) was mainly introduced as a system intended to target transactions involving less turnover, but still requiring notification in three or more Member States ("multiple filings"). Article 1(3) provides:

For the purposes of this Regulation, a concentration that does not meet the thresholds laid down in paragraph 2 has a Community dimension where:

(a) the combined aggregate worldwide turnover of all the undertakings concerned is more than ECU 2 500 million;

(b) in each of at least three Member States, the combined aggregate turnover of all the undertakings concerned is more than ECU 100 million;

(c) in each of at least three Member States included for the purpose of point (b), the aggregate turnover of each of at least two of the undertakings concerned is more than ECU 25 million; and

(d) the aggregate Community-wide turnover of each of at least two of the undertakings concerned is more than ECU 100 million;

unless each of the undertakings concerned achieves more than two-thirds of its aggregate Community-wide turnover within one and the same Member State.

10. As can be seen from the above, Article 1(3) employs a test of whether significant turnover is achieved in three Member States [2]. In principle, this means that the Article applies a test of qualified cross-border effects. In the discussions that led to the adoption of this amendment it was felt that the Community interest would be more manifest in cases involving three or more Member States and that, as long as only two Member States were involved, potential conflicts could be avoided through bilateral contacts. As will be seen later in this report, the focus on three, rather than two, Member States can be criticised as not targeting all transactions with a significant cross-border effect. In this respect Article 1(3) could be described as intended to provide companies recourse to the "one-stop shop principle" in situations where the cost of multiple notifications would otherwise be too high.

[2] It may also be noted that the current EEA-rules do not allow for Commission jurisdiction in cases where the Article 1(3) thresholds are met in a "mixed" configuration of Member States and EFTA-states. Thus, a concentration, which meets the Article 1(3) criteria in, for example, Denmark, Sweden and Norway, will not come under the Commission's jurisdiction.

11. This report will assess the operation of Article 1(2) and 1(3) thresholds from the perspective of the experience gained by the Commission, the NCA's and, finally from the viewpoint of the European business community. The data used for this report is primarily based on data relating to the period from March 1998 until December 1999 ("the reporting period").

I. Community dimension - the Commission's perspective

12. The total number of merger cases notified to the Commission will in general depend on three main factors:

a) the definition of the concept of a "concentration" in the Merger Regulation;

b) the level of turnover necessary to define "Community dimension"; and

c) the level of merger activity within the business community.

a) the concept of a "concentration" in the Merger Regulation

13. One of the amendments introduced in the Merger Regulation in 1998 was the inclusion of certain joint ventures, which previously had been considered to fall outside the scope of the Regulation (so called full-function co-operative joint ventures "ffCJV's").

14. In the reporting period, approximately 30 ffCJV's were notified. Out of these, only one or 3% was based on the turnover criteria of Article 1(3).

15. The extension of the Merger Regulation to include all ffCJV's is generally considered as appropriate. Experience has shown that these cases are well suited for the more structural type of assessment of the Merger Regulation. Their inclusion under the Merger Regulation has reduced the cost and delays involved in achieving regulatory clearance for the companies involved in such transactions. For the purposes of this report it is necessary to recall that the Commission's white paper on Modernisation (see above) foresees the possibility of a further expansion of the concept of a concentration, to incorporate also other joint venture cases, including those limited to production. Depending on the scope of the finally agreed extension of the concept of a concentration, this may be estimated to increase the number of cases notified under the Merger Regulation by at least 25-35 cases annually.

b) The level of turnover necessary to define "Community dimension"

16. If one for a moment disregards the dynamic development of the concept of a "concentration" described above, the number of cases notified to the Commission will be decided by the level of merger activity (see below), and the proportion of such mergers that will satisfy the existing turnover thresholds.

17. As an introductory point it may be noticed that the Merger Regulation, being based on mechanical turnover thresholds to divide competence between the Commission and the NCA's, includes two corrective mechanisms. The first such mechanism is the so-called 2/3-rule contained in both Articles 1(2) and 1(3). The operation of this rule will be assessed separately below.

18. The second corrective mechanism is the system whereby, on the one hand, one or more Member States can request the Commission to assess mergers that fall below the thresholds of the Regulation (Article 22). In the reverse, a Member State may, in cases that fall to be assessed under the Merger Regulation, request the transfer of competence if the transaction specifically affects competition in that Member State (Article 9). Although this referral system is not, as such, the object of this report, it may nevertheless be necessary to review the functioning of the referral process between the Commission and Member States should the existing levels of the turnover thresholds be revised.

Article 1(2)

19. The amendments to the Merger Regulation in 1998 did not affect the original turnover criteria in Article 1(2) [3]. The Commission's experience is that the criteria in Article 1(2) remain effective in identifying mergers that have a real Community dimension as opposed to a national dimension. As will be indicated below, the survey of representatives of the European business community has provided no indications that the applicable turnover thresholds in Article 1(2) should be raised.

[3] It may be noted that the absolute levels of the applicable turnover criteria are affected by factors such as inflation and fluctuations in the exchange rates. As an example, the original criteria of Article 1(2)(a), i.e. EUR 5 billion, corresponds to approximately EUR 4,3 billion when assuming an average 1.5% yearly inflation rate since 1990. Under the same assumption, the criteria of Article 1(2)(b), i.e. EUR 250 million, corresponds to approximately EUR 215 million.

Article 1(3)

20. In the reporting period, the Commission received 45 notifications under Article 1(3). This represented 9% of all notifications received in the period.

21. With a view to assessing the operation of Article 1(3), this section will provide a statistical comparison between cases notified under that Article and under Article 1(2). The comparison will focus, first on whether or not cases that are notified under Article 1(3) are more likely to be of a more national character. Thereafter, a comparison will be made as to whether cases notified under the two Articles present differences in their likelihood to raise competition concerns.

22. As far as the geographic area affected by the cases that were notified under Article 1(3) is concerned, one way to assess their impact is to look at the geographic market(s) affected by these mergers. In this exercise it first has to be stressed that the Commission often leaves the question of geographic definition open if the merger in question does not raise any competition concern under any of the conceivable alternate market definitions. Nevertheless, it is instructive that only five of the cases that were notified under Article 1(3) involved national geographic markets for all the relevant products [4]. Moreover, all of these five cases, which represented just below 13% of all cases notified under Article 1(3), affected more than one national market. Consequently, based on the Commission's experience there are no indications that cases notified under Article 1(3) affect a limited geographic area. As a point of reference it may be mentioned that about 12% of the cases notified under Article 1(2) were found to affect markets that remained national.

[4] ICI/Williams, case COMP/M.1167; Vedior/Select Appointments, case COMP/M.1702, AKZO Nobel/Hoechst Roussel Vet, case COMP/M.1681; Johnson and Johnson/Depuy, case COMP/M.1286 and Getronics/Wang, case COMP/M.1561

23. The decisions resulting from the cases notified under Article 1(3) show that at least half of these cases affected a relevant geographic market, which was as wide or wider than the EU. More specifically, in eight cases the affected geographic markets was at least as the EU as a whole [5], another ten cases involved geographic markets that were at least EEA-wide [6] and in one case the geographic market was world-wide [7].

[5] Lucchini/Ascometal, case COMP/M.1567; Getronics/Wang, case COMP/M.1561; Newell/Rubbermaid, case COMP/M.1355; Voest Alpine Stahl/Vossloh/VAE, case COMP/M.1259; Siebe/Eurotherm, case COMP/M.1195; UPM-Kymmene/April, case COMP/M.1006; Hyundai Electronics/LG Semicon, case COMP/M.1492; Huhtamaki Oyj/Packaging Industries Van Leer, case COMP/M.1656)

[6] Lucent Technologies/Ascend Communications, case COMP/M.1440; Dana/Glacier Vandervell, case COMP/M.1335; Constructor/Dexion, case COMP/M.1318; ELF Atochem/Atohaas, case COMP/M.1158; Ispat/Unimetal, case COMP/M.1509; Norsk Hydro/SAGA, case COMP/M.1573; Suez Lyonnaise/Nalco, case COMP/M.1631; Dupont/Sabanci, case COMP/M.1538; Dupont/Teijin, case COMP/M.1599 and Solutia/Viking Resins, case COMP/M.1763.

[7] Solvay/BASF, case COMP/M.1469

24. From a substantive viewpoint, it may be noted that competition concerns were found in four of the cases notified under Article 1(3). Three of these cases were cleared in the first phase by way of a decision under Article 6(1)b with commitments [8]. One case was cleared by way of an Article 8(2) decision with commitments, after a full second-phase investigation [9]. Consequently, over the period in question, just below 9% of the transactions notified under Article 1(3) raised competition concerns. This is exactly the same as the corresponding figure for cases that were notified under Article 1(2), where competition concerns were also found in approximately 9% of the cases.

[8] Akzo Nobel/Hoeschst Roussel Vet, case COMP/M.1681; Pakhoed/Van Ommeren II, case COMP/M.1621 and Johnson&Johnson/Depuy, case COMP/M.1286

[9] Sanitec/Sphinx, case COMP/M.1578

25. Based on the available information, it appears that the mergers notified under Article 1(3) generally affect markets that are wider than national. It appears, therefore, that the mergers identified by Article 1(3) have a clear Community dimension. In addition, these cases appear as likely to raise competition concerns as those notified under Article 1(2).

The 2/3-rule

26. It is difficult to find reliable statistics on the application of the 2/3 rule in cases that did not meet the turnover thresholds in Article 1. The reason for this is that the national notification requirements applicable in the Member States do not oblige parties to specify their EU turnover for each Member State. Still, some Member States have been able to provide estimates as to the number of cases submitted to them that fulfilled the 2/3 requirement. These estimates are set out in the next section below.

27. As to the Commission's own experience, this consists basically of its knowledge that certain specific cases that would otherwise have been notified to it failed to meet the requirements in Article 1 owing to the 2/3 rule. The Commission's information in this respect is by no means complete. A recent example of a case with significant cross-border impact that failed to meet the Article 1 requirements owing to the 2/3 rule was the concentration between the Chase Manhattan Corporation and Robert Flemmings Holdings Limited. Both of these international financial companies manage assets that are measure in hundreds of billions USD and are active in 40-50 countries world-wide. Still, both companies have more than 2/3 of their combined turnover in the United Kingdom.

28. Another example of how the 2/3 rule refers similar transactions to different jurisdictions is the two recent big mergers in the German electricity market, namely VEBA/VIAG and RWE/VEW. Both transactions involve large operations in Germany and, to a smaller extent, in other Member States. Although both operations have their immediate impact on the electricity markets in Germany, a significant impact on the electricity markets in neighbouring countries was likely to occur. In the RWE/VEW case, each of the parties achieved more than 2/3 of their aggregate turnover in the same Member State. In the VEBA/VIAG case, only one party achieved 2/3 of its aggregate turnover in one Member State, and the turnover of the other party was just below the 2/3 level. Consequently, the VEBA/VIAG case fell under the Commission's jurisdiction, whereas the RWE/WEW case did not.

c) The level of merger activity within the business community.

29. The below table illustrates the development in terms of number of notified cases under the Merger Regulation over the last 10 years.

>TABLE POSITION>

30. The above statistics serve to illustrate the continuing "merger-boom", which in itself has many explanations (completion of the internal market, introduction of the Euro, globalisation of the economy, preparation for enlargement of the Community etc.).

31. In relation to this particular report on the operation of the thresholds of Article 1 of the Merger Regulation, the above statistics are interesting in that they show that the increase in notified cases is not caused by the amendments to the Regulation in 1998. This is illustrated by the increase from 1998 to 1999. The available statistics for the first four months of 2000 does not indicate a reverse in this trend (95 notifications, compared to 92 in the same period of 1999). It should be stressed that there are no signs that the high level in merger activity will not continue into the foreseeable future. Given the obvious impact on Commission resources of this development, the feasibility of any suggestion that would have the effect of further increasing the number of cases notified under the Merger Regulation would have to be carefully analysed.

d) Conclusion

32. In conclusion, based on the Commission's experience so far, the turnover criteria of both Articles 1(2) and 1(3) are well suited to catch mergers with a Community dimension. Consequently, there appear to be no objective reason to consider any measures that would limit the Commission's jurisdiction as concerns such transactions.

33. The fact that most of the mergers that fulfil the existing turnover requirements have significant cross-border effects cannot, however, provide an answer to the question whether the existing thresholds catch all (or even most) mergers with a Community dimension. This latter question will be assessed below from the perspective of the NCA's as well as from the viewpoint of the business community.

II. Community dimension - the NCA's experience

34. In the context of the ongoing co-operation between the Commission and the NCA's, statistics on concentrations notified in accordance with national laws have been collected for the period from March 1998 to December 1999. National notification requirement differ as far as the amount of requested information is concerned, e.g. some, but not all, include information on the merging parties' EU-wide turnover, whether multiple filings were required, and so on. Still the information supplied to the Commission is sufficient to draw a number of preliminary conclusions on the operation of the turnover thresholds in Article 1.

Multiple filings

35. One way of analysing the operation of the turnover thresholds is to look at the number of cases which, despite the introduction of Article 1(3) has had to be notified in several Member States. The available statistics [10] indicate that in approximately 8%, or 364 out of 4303 transactions, the notifying parties submitted their notification to more than one NCA [11]. Logically, that the national notification requirements in each Member State must be assumed to be intended to catch transactions that typically can have a significant impact in that Member State. It therefore follows that the fact that these 364 concentrations had to be notified in more than one Member State provides, in itself, an indication that these transactions must have had a significant cross-border impact.

[10] Information from Greece is still awaited.

[11] With regard to the number of undertakings involved in the concentrations, the large majority (84%) involved two undertakings, in 10% of the cases there were three undertakings involved, and, finally, in 4% of the cases there were more than three undertakings involved. These figures have remained constant over the period of data collection.

36. Further analysis show that out of the 364 transactions that were notified in more than one Member State, the majority of these cases were notified in two Member States (294 transactions). Relatively few concentrations were notified in three Member States (31 transactions) and 39 transactions were notified in more than three Member States [12].

[12] The geographic aspect of the multiple-filing cases, i.e. which Member States were involved, have not been further evaluated, as it is not of immediate importance for this report.

37. Considering that Article 1(3) applies a qualified test of cross-border effects, where significant activities in three Member States is required, it can be argued that the transactions that had to be notified to two NCA's should not raise any concerns. This conclusion would, however, not remain valid, should it be considered appropriate to remove the three-country qualification requirement from Article 1(3).

38. On that basis, less than 2% of all transactions notified to the NCA's were of the type that Article 1(3) is intended to catch, i.e. those that had to go through more than two national notifications. Still, it cannot be ignored that the number of transactions that were notified to three or more NCA's (70 cases) clearly exceeded the number of cases notified to the Commission in accordance with Article 1(3) under the same period (45 cases). In conclusion, the information received from the NCA's indicates that Article 1(3) in its current form has not removed the need for "multiple filings" of a significant number of transactions. In order to answer the question why these transactions failed to meet the requirements of Article 1(3), a more detailed analysis of the circumstances surrounding those particular cases would be needed.

Analysis of turnover levels in transactions notified to the NCA's

39. The Commission has asked the NCA's to provide turnover figures for the cases notified during the reporting period. The requested information relates to the world-wide turnover, EU turnover, and the turnover achieved in individual Member States. It should, however, be noted that the level of detail in the turnover information that notifying parties are required to submit under national law varies among Member States. The information used for this report is that which was available to the NCA's. This means that there is an uncertainty as to how the turnover information relates to the Merger Regulation requirements. Importantly, to the extent that Member States do not use the same group concept as in Article 5, there will be a discrepancy in the submitted statistics. Thus, even if the currently available information can be used to show certain trends, more detailed information would be desirable prior to drawing firm conclusions.

40. First, the turnover information relating to all concentrations notified in more than one Member State, i.e. the 364 cases notified to two or more NCA's have been analysed with regard to the thresholds set out in Article 1(2) of the Merger Regulation. For the whole period March 1998 to December 1999, about 25% of the notified transactions had a combined world-wide turnover that exceeded the EUR 5 billion set out in paragraph 1(2)(a). These cases did, however, not meet the criteria of paragraph 1(2)(b), i.e. the aggregate Community-wide turnover of one of the undertakings concerned was below the threshold of EUR 250 million. Notably, in many of these cases (85%), the Community-wide turnover of the target company was less than EUR 150 million. Furthermore, on the basis of estimates provided by some NCA's, the two-thirds rule applied in about 70% of the transactions that met the EUR 5 billion threshold.

41. Given the relative novelty of Article 1(3), special emphasis has been given to the application of the thresholds set out in that Article. As mentioned above, during the reporting period, 45 cases were notified to the Commission under the criteria set out in Article 1(3). This represents approximately 9% of all notifications (494) received by the Commission during the same period.

42. The information provided by the NCA's has been analysed with regard to the thresholds set out in Article 1(3) of the Merger Regulation. For the period March 1998 to December 1999, 36% of the notified transactions exceeded the combined world-wide turnover of EUR 2.5 billion set out in Article 1(3)(a).

43. A number of these transactions failed to meet the three country requirements set out in paragraphs (b) and (c) of Article 1(3). In a certain sense these requirements have to be seen in context. Together, they are intended to ascertain that transactions notified to the Commission involve a minimum level of aggregate turnover (EUR 100 million) and individual turnover (EUR 25 million) in at least three Member States. Most of the cases that did not qualify for Community dimension failed one or both of these tests. In fact, the criteria of Article 1(3)(c) appears to have been decisive in the greatest proportion of cases, meaning that the EUR 25 million threshold was not met in a third, required, Member State. However, the available statistics could only indicate a very approximate number for how many of the multiple filing cases that where excluded from Community jurisdiction by virtue of Article 1(3) (b) and (c) respectively. Thus, even if the currently available information can be used to show certain trends, more detailed information would be desirable prior to drawing firm conclusions.

44. Furthermore, a number of the transactions did not meet the individual Community-wide turnover requirement of EUR 100 million, as set out in paragraph (d) of Article 1(3). Moreover, on the basis of estimates provided by some NCA's, the two-thirds rule applied in about 65% of the transactions that met the EUR 2.5 billion threshold. However, also here the available statistics could only indicate a very approximate number for how many of the multiple filing cases that where excluded from Community jurisdiction by virtue of Article 1(3) (d).

45. The available turnover figures allows a tentative conclusion to be drawn, indicating that some 25-30 cases (of the 70 that were notified in three or more Member States) were close to satisfy the criteria set out in paragraph (b) of Article 1(3). For these transactions, the available information does not, however, allow any firm conclusions to be drawn as regards the criteria of paragraphs (c) and (d). Thus, even assuming that factors such as inflation could, over time, allow some cases that are already "close" to eventually come into the scope of the Merger Regulation, less than half of the cases that were notified in three or more Member States actually came "close". Moreover, on the same basis, less than 10% of the cases that were notified to two or more Member States came "close" to the scope of the Merger Regulation.

46. In order to allow better comparison of the merger statistics received from Member States, an overview of the statistical analysis over the enquiry periods is set out in the table below.

>TABLE POSITION>

47. In addition to the collection of statistics from Member States, the Commission has collected some statistics from transactions that were discussed at a pre-notification stage, but which finally failed to reach the thresholds pursuant to Article 1(2) and/or Article 1(3). Whilst the number of such cases is quite limited (14), the result appears to be broadly in line with that of the experience from the NCA's. In most of these 14 cases, the world-wide threshold according to Article 1(2)(a) and Article 1(3)(a) was met. Three of the cases failed to meet the aggregate Community-wide turnover levels required in Article 1(2)(b) and three cases failed to reach the aggregate Community-wide turnover required by Article 1(3)(d). Eight operations failed to meet the criteria pursuant to Article 1(3)(c) as the EUR 25 million threshold was not met in more than two Member States.

Geographic impact of the transactions notified to the NCA's

48. As indicated above, in the section on the Commission's experience from Article 1(3) cases, one way to measure the Community impact of a given concentration is to see whether the affected market is wider than national (or whether more than one national market is affected). The number of cases notified to the NCA's is significant (more than 4000 cases). It has therefore not been possible within the time available for the preparation of this report to analyse to what extent the cases handled by the NCA's could properly be defined as affecting more than one national market. However, prior to making any suggestion to amend the existing thresholds, it would clearly be preferable to conduct such an analysis, at least in relation to the cases that had to be notified in more then one Member State.

III. Community dimension - perspective of the European industry

Introduction

49. As part of the preparation of this report, the Commission has invited firms and their representatives (business associations and law firms) to respond to a questionnaire. The questionnaire was primarily designed to collect further information concerning the application of the thresholds in Article 1, but also on matters related to the application of the thresholds. Furthermore, the respondents were invited to forward any comments or suggestions in relation to further improving the functioning of the Merger Regulation.

50. First, the answers received from the business community as regards the existing thresholds will be presented from a statistical perspective. Second, an overview will be provided to indicate the most prevalent qualitative views regarding the need for further review.

Analysis of the replies related to the operation of the thresholds

General statistics on replies

51. Table 1 below shows from which type of organisations replies were made.

>TABLE POSITION>

Statistics on the individual questions

52. Table 2 below outlines the replies given on a question by question basis and further divided according to the type of organisation. The text of the questions is repeated in short form.

>TABLE POSITION>

Qualitative replies to the questions related to the operation of the thresholds

53. Most respondents have stressed that, seen from the business community's viewpoint, the main objective of the Merger Regulation is to provide fast, efficient and foreseeable decisions on the competitive impact of mergers and similar transactions. Many of the respondents have made reference to the "one-stop shop principle" as well as the necessity for a "level playing field". There is a general view that the current turnover thresholds of Article 1 do not allow the business community the full benefit of these guiding principles (in the sense that a substantial number of deals with significant cross-border effects remain subject to multiple national control). For this reason most respondents are in favour of a reduction and simplification of the thresholds, as well as of any other proposal that would lead to a consistent application of the one-stop shop principle.

Comments regarding the Article 1(2) thresholds

54. As can be seen from the table above, most respondents are of the view that the Art. 1(2) thresholds are appropriate to confer jurisdiction to the Commission (78%) and that these thresholds appropriately indicate a Community-dimension (65%). As none of the respondents have argued for an increase in the applicable thresholds, it appears that the less than 100% rate of satisfaction is due to the prevalent feeling that the thresholds should be lowered. Consequently, many respondents have submitted that several factors, such as the completion of the internal market, have as an effect that a significant proportion of concentrations below these levels has effects beyond a single Member State.

55. In this context, many respondents have suggested that that the current turnover thresholds should be reduced, thus enlarging the Commission's jurisdiction in merger cases. The most common suggestion is that the combined world-wide turnover requirement should be lowered to EUR 2 billion, whereas the Community turnover requirement should be set at EUR 100 million for two parties. (None of the respondents have provided any "scientific" support for the introduction of these particular levels. Instead, they often refer to the Commission's statement in the notes to the Merger Regulation at the time of its adoption in 1989, where the Commission considered that a lowering to EUR 2 billion would be appropriate.) Many of the respondent who have made this proposal have indicated that such an amendment would allow Article 1(3), which they consider to be complicated and not sufficiently effective, to be deleted.

56. A number of other suggestions aimed at lowering the thresholds have been put forward by respondents. One such suggestion is that the requirement relating to world-wide turnover should be removed in favour of an increased focus on EU/EEA turnover (on the basis that the world-wide criteria is an imprecise measure of Community impact). Other suggestions include moving away from a turnover-based measurement and to introduce a measurement based on assets or market share at the European level.

57. As ancillary suggestions to a significant increase in the Commission's competence to assess mergers, it has been suggested that a more restrictive approach as regards the level of required information in a notification should be adopted for transactions that are unlikely to have negative effects on competition. It has also been suggested that the stand-still provisions should not apply to such transactions.

Comments regarding the Article 1(3) thresholds

58. A significantly lower proportion of the respondents find that the Art. 1(3) thresholds are appropriate to confer jurisdiction to the Commission (62%). Moreover, a majority of the respondents have replied negatively to the question whether the Article 1(3) thresholds are appropriate to determine that a concentration has a Community-dimension (54%). The comments that have been submitted do not question that transactions that meet these thresholds have a Community dimension. On the contrary, the main criticism is that the Article 1(3) thresholds excludes a significant number of transactions from the advantages of the one-stop shop principle and thereby leads to too many multiple national filings. Some respondents have indicated that the requirement of significant turnover in three, rather than in two, Member States considerably reduces the applicability of Article 1(3) to transactions with significant cross-border effects. This view is largely supported by the statistics submitted by the NCA's, see section II above.

59. The responses also indicate that the criteria in Article 1(3) are seen as overly complex (compared to those in Article 1(2)). In particular, many respondents have indicated that the necessity to identify the relevant turnover figures for each Member State can be both costly and time consuming. These costs often have to be borne by firms simply to find out that a particular transaction is not notifiable under the Merger Regulation. In such cases, the involved firms will obviously also need to bear the costs associated with national filing requirements. The difficulty of calculating the necessary turnover figure have been indicated to be particularly high in emerging markets (such as the information technology markets), where, in addition, it has been stated that the strong expansion tend to make the last available audited turnover figures an unreliable measure of Community interest.

60. In conclusion, the respondents believe that the one-stop-shop principle should also apply to smaller operations with significant cross-border effects. The main difficulties that they see in multiple filings is the creation of significantly higher costs, e.g. for local counsel and translations, and that multiple filings are more time consuming as they require screening of many national legislation systems. Another perceived problem is that multiple filings could lead to contradictory results for the transaction as national legislation differs in the Member States and the ambiguous timing constraints for the implementation that results from being subject to several national merger control regimes. Finally, the number of NCA's involved is perceived to multiply the complications in discussing remedies to identified competition concerns.

61. In view of these criticisms many respondents have suggested that Article 1(3) should be repealed and that the thresholds in Article 1(2) should be substantially lowered. (As indicated above, the most common suggestion is that the combined world-wide turnover requirement should be lowered to EUR 2 billion, whereas the Community turnover requirement should be set at EUR 100 million for two parties.)

62. Another proposal, sometimes referred to as "second best", is that the turnover levels in Article 1(3) should be lowered and/or that a "two-country requirement" should replace the current "three-country requirement". The most common proposal for lowering these thresholds is that the EUR 2 billion should be maintained for the world-wide threshold, whereas the "three-country" thresholds should be lowered. More specifically, respondents have proposed that the combined EU-wide threshold (Article 1(3)b) should be lowered to EUR 50 million and that the individual turnover threshold (Article 1(3)c) should be lowered to EUR 15 million.

63. Some respondents have also suggested that the Commission should automatically have jurisdiction in cases where notification would otherwise be necessary to three or more NCA's. (As indicated above, this was the Commission's original proposal in the 1996 Green Paper. It was, however, rejected in the course of the consultations as being too complex.)

Comments regarding the 2/3 rule

64. The respondents were also invited to express their views as to whether the 2/3 rule (contained in both Articles 1(2) and 1(3)) reflect satisfactorily operations with a national dimension. Whereas a majority (65%) of the respondents have replied that this is the case, a large proportion of the respondents believe that many transactions that have a significant cross-border impact are excluded from the one-stop shop principle owing to the 2/3 rule.

65. In particular, it has been submitted that 1/3 of the turnover could already reflect very significant activities abroad, in particular when larger firms are involved (as is intuitively the case under the current EUR 5 billion threshold). Respondents have also stated that this could lead to an unjustified discrimination between firms that have their core businesses in larger Member States (more likely to fulfil the 2/3 rule) compared to firms having their core business in a smaller Member State.

66. Various suggestions have been proposed to reduce the impact of this rule, by increasing the proportion of turnover required to be achieved in a single Member State. These proposals vary from 3/4 to 5/6 of the relevant turnovers. Other respondents have submitted that the 2/3 rule is not necessary, as the instrument of Article 9 will be sufficient to safeguard any legitimate specific concerns of individual Member States. However, at the same time a number of respondents have sated that an excessive use of Article 9 decreases legal certainty and that the scope of this Article should be strictly interpreted.

Comments relating to issues outside Article 1

67. In addition to the questions relating to the turnover thresholds in Article 1, respondents were also invited to provide general or specific comments aiming at improving the merger control system. A wide-ranging spectrum of comments has been received. Some comments concern the referral provisions in Articles 9 and 22. Given the close relationship between the referral provisions and the turnover thresholds these will be set out below [13].

[13] Other issues raised by the business community relate to other substantive provisions of the Merger Regulation, as well as procedural and institutional issues relevant to the Commission and to the NCA's. These are set out in the annex to this Report.

Referral provisions

68. As has already been indicated above, a number of respondents are sceptical to an increased use of Article 9 (referral from the Commission to a Member State). These respondents are of the opinion that referrals may create additional costs and increase the legal uncertainty. Reasons for additional costs may be that it is necessary to consult with legal experts in the Member State in question, translations and possibly longer waiting periods. The increased uncertainty has been indicated as being caused by the fact that merger control in a number of Member States, in addition to competition criteria, rely on various expressions of "public interest". Some respondents have also mentioned the fact that several national merger control systems, at various stages, call for the possible intervention on a political level.

69. As regards Article 22 (referral from one or more Member States to the Commission), some respondents have stated that measures should be adopted to allow effective use of this provision in multiple filing cases. Although the 1998 amendments to the Merger Regulation introduced such a possibility, there have still not been any joint referrals by two or more Member States to the Commission under Article 22(3). The information available at this stage does not allow any conclusions as to the reasons why joint referrals have not been made in the cases where a transaction has undergone multiple national notifications.

IV. Preliminary analysis and impact on resources

70. As described above, the Commission's experience shows that concentrations, which meet the turnover thresholds in Article 1(2) and/or Article 1(3), generally have a significant cross-border impact. This finding is confirmed by the views expressed by the European business community (which, in addition, would largely favour an increase in the number of cases handled by the Commission). In contrast, a number of Member States have expressed some doubts as to the need to change the current jurisdictional division between the Commission and the Member States. Nevertheless, most Member States agree that there is a need for a more in-depth analysis of these issues, and such an analysis should be based not only on a statistical approach, but also on an assessment of the effects of multiple filings [14]. The Member States have also expressed their continued willingness to provide the Commission with further empirical data.

[14] Member States have also raised a number of issues that they consider relevant for a more in-depth review. These are set out in the annex to this Report.

71. There are, however, in the Commission's view several indications that the current thresholds have as an effect that an important number of cases that have a significant cross-border impact fall outside the Community rules. The figures received from the NCA's strongly suggest that a lowering of the overall world-wide turnover requirement is not the key to this problem. In fact, already at the current levels, 25% of all notifications to the NCA's met the EUR 5 billion threshold in Article 1(2) [15]. All these cases either failed to meet to EUR 250 million threshold or were caught by the 2/3 rule. At the same time 36% of the notifications to the NCA's exceeded the EUR 2.5 billion threshold in Article 1(3). Here, the available information suggests that about 10% (or approximately 400 cases) of these cases also exceeded the EUR 100 million threshold in 1(3) (d). Consequently, these cases failed either one of the three Member State requirements or the 2/3 rule. The available statistics are not sufficiently detailed to allow a more detailed conclusion as to the most prevalent "stumbling block".

[15] The percentage figures indicated here are based on the turnover figures that the NCA's have provided to the Commission, i.e. in relation to multiple filing cases. Further data collection would be needed for an assessment of the profile of the remaining cases dealt with by the NCA's.

72. Despite the above reservations, the available information indicates that the two main reasons why these cases fail to meet the existing thresholds are, first, the requirement in Article 1(3) of a minimum amounts of turnover in three Member States, and, second, the 2/3-rule. (A third reason could be that, inter alia, certain "new technology" industries generally do not generate turnover at the levels required by the Merger Regulation.)

73. The available data allows some cautious quantification of the effects of the current criteria of Article 1(3). What appears to be clear from the statistics provided by the NCA's is that the level of turnover required is less important in excluding Community jurisdiction that the number of Member States, where such turnover in required. This means that even if the level of required turnover would have been lower, it is still likely that "only" the 70 cases that where notified to three or more Member States would have been candidates for Article 1(3). The fact that this figure of 70 cases is significantly higher than that notified according to Article 1(3) (45 cases), provides a strong indication that the current criteria should be reviewed. It should also be noted that the available information does not allow conclusions as to by how much the Article 1(3) thresholds would need to be reduced to bring the 70 cases within the scope of the Merger Regulation.

74. On the other hand, a reduction of the number of "affected" Member States, from three to two, would have a higher potential to make some or all of the 364 cases that were notified to more than one NCA eligible for treatment under the Community rules.

75. As indicated in the introduction, the Merger Regulation was originally intended to provide a level regulatory playing field for all mergers that could be presumed to have significant cross-border effects. The available information provides strong indications that the current thresholds of Article 1 are less than optimal for this purpose and do not, for many transactions with significant cross-border effects, provide a one-stop shop merger control. In fact, certain transactions have even after the introduction of Article 1(3) had to be notified to seven or more Member States (as well as to several NCA's in applicant countries) [16].

[16] For example, CSM/Leaf, notified in nine Member States (B, D, DK, Ir, I, NL, S, SF & UK), United Technologies Corp/Electrolux Commercial Refrigeration, notified in eight Member States (A, D, DK, H, NL, I, S & SF), as well as in Croatia, Estonia, Poland and Romania.

76. As far as the impact of the 2/3-rule is concerned, it is the Commission's experience that certain transactions with clear cross-border effects are excluded from the scope of the Merger Regulation by this criterion. It may also be noted that a significant proportion (35%) of the European business community representatives has indicated that this rule does not distinguish "national transactions" in a satisfactory manner. Still, it has to be concluded that the available information is insufficient to allow any firm conclusions.

77. Consequently, further data collection and analysis would be required prior to making any concrete conclusion or proposing specific amendments. Nevertheless, even the available figures indicate that the introduction of a consistent approach, where all transactions with a significant cross-border effect were to be treated under the Community rules would have a significant impact on the Commission's resources [17].

[17] In principle, an alternative possibility would be to further harmonise the national rules, possibly in combination with a more flexible referral system. However, on the basis of the comments received so far from the business community, this does not appear to be favoured by them, as any deviation from the one-stop shop principle is perceived to decrease predictability and increasing the regulatory costs. However, the justification for such claims should be further analysed in the course of the wider review.

78. The above-mentioned impact on resources can be demonstrated by reference to the number of cross-border transactions handled by the NCA's in the reporting period. As indicated before, this was at least 364 cases. Compared to the total number of cases notified to the Commission during the same period (494 cases), this clearly is a significant number of cases. There are also reasons to assume that the actual number of cases notified to the NCA's with significant cross-border effects is even higher. This is due, partly, to the fact that the national systems for merger control are not yet harmonised, for example, some Member States have no merger control, and others are based on voluntary notification [18]. This most likely has the effect that a certain number of cases with significant cross-border impact do not enter the statistics of cases notified to more than one NCA. In addition, if cross-border impact were to be measured on the merits, e.g. by whether the relevant markets are wider than national (or whether more than one national market is affected), it is likely that an additional number of cases would fall to be assessed under the Community rules. Moreover, the proportion of such cases will increase with the completion of the internal market. [19]

[18] Of the Member States with voluntary notifications systems, Spain introduced a mandatory regime in 1999, and proposals for such a change is pending in France and in Denmark. Also the UK is considering certain changes to its merger control rules. The introduction of additional national control is widely expected to further increase the number of multiple notifications.

[19] An additional increase in the number of notifications under the Merger Regulation will result if the proposal in the white paper on Modernisation - to treat additional joint ventures under the Merger Regulation - is adopted.

79. These potential effects on the Commission's resources are widely recognised by the business community. Many respondents have observed that there has been a large increase in the number of transactions being notified under the Merger Regulation over the last years. These respondents generally expect this trend to further accentuate in the years to come. In addition, as has been explained above, the respondents are generally in favour of substantially lowering the thresholds in Article 1. Unless the necessary resources for handling a significantly increased number of merger cases can be found by other means, it has been suggested that the Commission could introduce a filing fee for merger notifications [20].

[20] Filing fees are currently applied in some jurisdictions, including the US, the UK, Germany, Austria and Spain. At the European level, a figure of EUR 30.000 has been mentioned as not being unreasonable. On the basis of, say, 400 notifications annually, such a system would generate EUR 12 million, which would allow a non-insignificant addition of human resources.

V. Conclusion

80. In conclusion, at this stage it appears that the criteria in Article 1(3) and perhaps even those in Article 1(2) have as a result that an important number of transactions with significant cross-border effects, and therefore a Community interest, remain outside the Community rules on merger control. There are also indications from the European business community that this is considered as problematic, and that it acts as a constraint on the investment decisions of European firms.

81. On that basis the Commission believes that a more in-depth analysis of the appropriate mechanism for establishing Community jurisdiction in merger cases is necessary. At this stage it would also appear that any change to the Merger Regulation that would remove this unbalance is likely to require:

* To significantly change the existing system of case allocation (work distribution) between the Commission and the NCA's.

* The attribution of significant additional Commission resources dedicated to treating all mergers with significant European cross-border effects.

* A more thorough review, not only of the existing turnover thresholds, but also of other substantive and procedural rules relating to the control of concentrations.

82. The Commission therefore invites the Council to take note of the information set out in this Report and to endorse the further investigation, in conjunction with the Member States and other interested parties, into the appropriate mechanism for establishing Community jurisdiction in merger cases as well as other matters related thereto.

83. The Commission will also submit this report for information to the European Parliament and the Economic and Social Committee.

ANNEX TO THE COMMISSION'S REPORT TO THE COUNCIL ON THE APPLICATION OF THE MERGER REGULATION THRESHOLDS

1. The Commission has received comments from Member States as well as from the business community about the desirability to include issues outside the scope of Article 1 in the review of the Merger Regulation. These proposals have been summarised in this annex, which, however, does not constitute an exhaustive list of all proposals received.

2. In addition, it is possible that this list of non-threshold issues may be expanded, in particular since the more in-depth analysis is likely to involve contacts with interested parties who have not yet made their views known. Furthermore, the Commission will necessarily have to pay particular attention to the need to maintain and develop the consistency of the European merger control system as a whole. It cannot be excluded that this will have to involve proposals, including certain "housekeeping amendments", and that this may go beyond the matters mentioned in this Report.

A. "Non-threshold issues" raised by the Member States in relation to the possible review of the functioning of the Merger Regulation

3. A large number of Member States have proposed that the review should cover, in addition to the thresholds, the existing rules for referrals (Articles 9 and 22). Member States regard these provisions as important mechanisms for fine-tuning the effects of the turnover-based thresholds in Article 1. In general terms it has been proposed that the aim of the review should be to improve the efficiency and transparency of Articles 9 and 22. Moreover, several Member States have referred to the involved companies' need for legal certainty in relation to referrals, as well as to the usefulness of collecting objective information concerning the impact, if any, on the costs that involved firms will have to bear in referral cases.

4. Member States have also indicated the need to review the procedural rules relating to their role (through the Advisory Committee on concentrations) in merger cases where commitments are offered to remedy competition concerns. The Member States are concerned that the current system does not always allow for an effective and transparent discussion, due to the various time constraints in the procedure.

5. Finally, Member State representatives have proposed a number of other issues for the review. These include the possibility of introducing a system of streamlined treatment of "routine cases" into the Merger Regulation. Moreover, it has been suggested that effectiveness of the rules relating to fines and periodic penalty payments in Article 14 and 15 should be reviewed.

6. Several Member States have indicated that the review should consider the impact of enlargement of the European Union.

A. "Non-threshold issues" raised by the business community in relation to the possible review of the functioning of the Merger Regulation

7. The obligation to notify concentrations with a Community dimension is triggered by the conclusion of the agreement, or the announcement of the public bid, or the acquisition of a controlling interest (see Article 4(1)). These concepts have been further explained in the Commission's Notice on the concept of a concentration. Still, respondents have suggested that the concepts of "announcement", "agreement", "acquisition of control" should be further clarified. Some respondents have suggested that notification should be possible even prior to the conclusion of a legally binding agreement.

8. Other respondents have suggested that the rules in Article 5 relating to the definition of the turnover calculation should be revised and simplified, in particular as regards financial institutions.

Procedural and institutional issues

Resources

9. As indicated in the Report, many respondents from the business community would be in favour of substantially lowering the thresholds in Article 1. As a means of financing the increased number of notifications, the business community appears willing to accept the introduction of a filing fee for merger notifications.

10. In this context there is widespread concern that the Commission must allocate appropriate resources in order to cope properly with the expected workload of merger cases, so as to maintain the current level of openness and quality. Practical examples have been given of the negative consequences flowing from an insufficient level of appropriate resources, including the reduced time that Commission officials can devote to any particular case. Another example that has been mentioned relates to the inconvenience and loss of momentum that can be experienced should it for resource problems be necessary to change the members in the case team appointed to investigate a notification [21].

[21] Relatively simple measures of a technical nature, such as the introduction of a voice mail system, where a message could be left in case of absence of case handlers, have also been suggested as necessary priorities.

Prioritisation

11. Furthermore, it has been suggested that measures should be adopted that would better allow the Commission to focus its resources on notified transaction that will have a significant impact on competition. For these cases, some respondents have indicated that the one-month period of the phase 1 investigation can be too short to allow effective treatment of the case and to be transparent. Therefore, the possibility of having a slightly longer first phase investigation, for example six weeks, has been mentioned, in particular when complaints have to be treated. Also for other, simpler cases, it has been suggested that additional measures should be adopted to decrease the burden imposed on notifying parties (e.g. simplification of the Form CO, used for notifying concentrations to the Commission).

Procedural issues

12. Some respondents have proposed that there should be a legal deadline for the Commission's possibility to declare a submitted notification to be incomplete (such a declaration re-starts the one-month period, see Article 4(2) of the Commission's Implementing Regulation, EC No 447/98). Various proposals have been made as to the appropriate length of such a deadline (within 1-3 weeks after notification).

13. Some respondents have suggested that the existing rules on treatment of business secrets and other confidential information should be reviewed. Suggestions have been made in both directions. Some have suggested that a higher degree of protection should be imposed, for example in that the Commission should take further measures to make sure that sensitive information is not divulged in its publications, e.g. in press releases. Other respondents have suggested that the current degree of protection should be lowered, for example by allowing notifying parties access to complaints submitted by third parties at an earlier stage in the procedure than is currently foreseen by Article 18. (The notifying parties have access to the Commission's file after receiving a "Statement of Objections").

Procedure at Member State level

14. Some respondents have suggested that the merger control procedure at the national level should be further harmonised in order to reduce the costs and uncertainties associated with multiple filings. The proposed harmonisation measures include that the national systems should align their notification thresholds up to the level where cases have to be notified to the Commission. Other suggestions are that multiple notifications (to several NCA's) should be allowed to be made in one Community language and that the national statutory waiting periods should be harmonised.

Procedure at the level of the Court

15. The current system for appeal of Commission decisions under the Merger Regulation is regarded by some as unsatisfactory. It has therefore been suggested that a fast-track procedure at the EU courts should be established concerning the appeal of merger decisions.

Top