Important legal notice

61997C0044

European Court reports 1999 Page I-07177

Opinion of the Advocate-General

1 This action, brought before the Court of Justice by the Federal Republic of Germany, is for the partial annulment of Commission Decision 96/701/EC of 20 November 1996 amending Decision 96/311/EC on the clearance of accounts of the Member States for certain expenditure financed by the European Agricultural Guidance and Guarantee Fund (EAGGF), Guarantee Section, for the financial year 1992 and certain expenditure for the financial year 1993, (1) in so far as, by that decision, the Commission charged to the Federal Republic of Germany a total sum of DEM 19 591 000 (hereinafter `the amount charged').

2 That amount corresponds to a flat-rate correction of 2% of the total declared expenditure for the delivery of beef into intervention during the 1992 financial year.

3 The Commission applied that flat-rate correction in accordance with its guidelines on the `Financial consequences within the framework of the clearance of accounts of EAGGF Guarantee Section of deficiencies in controls carried out by the Member States - flat-rate corrections'. Those guidelines provide for flat-rate reductions of 2%, 5% and even 10%, depending on the seriousness of the deficiencies. The 2% reduction applied in this case, operates `where the deficiency is limited to parts of the control system of lesser importance, or to the operation of controls which are not essential to the assurance of the regularity of the expenditure, such that it can be reasonably concluded that the risk of loss to the EAGGF was minor.'

4 The Commission justifies applying that flat-rate reduction on the ground that it identified a number of deficiencies during the audit it conducted, in 1993 and 1994, of measures for the purchase and sale as well as the storage of beef and veal for intervention, at the Bundesanstalt für landwirtschaftliche Marktordnung (the German intervention agency; hereinafter `BALM') in relation to four coldstores, as well as on the basis of the annual inventories of the 107 coldstores in Germany.

5 More specifically, in its summary report on the results of inspections concerning the clearance of the EAGGF Guarantee Section Accounts for 1993, (2) the Commission listed a series of complaints concerning controls on entry into storage of the meat, during storage and on removal from storage.

6 According to the Commission, those deficiencies in the control system constitute infringements of Article 8(1) of Council Regulation (EEC) No 729/70 of 21 April 1970 on the financing of the common agricultural policy, (3) as well as Articles 3 and 4 of Commission Regulation (EEC) No 618/90 of 14 March 1990 laying down rules for drawing up the annual inventory of agricultural products in public storage. (4)

7 The Commission makes the point that its findings and the 2% flat-rate correction relate to the inadequacy of the controls as a whole, that is to say the deficiencies at all three stages (controls on takeover, during storage and on removal from storage) and not the inadequacies specific to each of the three stages. The Commission considers that it is the shortcomings in the German system of control generally that pose a risk for Community resources. Furthermore, the Commission has confined itself to a 2% correction because the Federal Republic of Germany has made an effort to remedy the deficiencies identified and spot checks have actually made it possible to uncover abuses.

I - The background to the dispute

8 During the period 1994 to 1995, views were exchanged and discussions held on a number of occasions, including on 20 January 1995 at BALM's Frankfurt office, between the competent German authorities and the Commission, but these proved fruitless. Subsequently, on 15 December 1995, the German Government requested that the conciliation procedure be initiated in accordance with Commission Decision 94/442/EC of 1 July 1994 setting up a conciliation procedure in the context of the clearance of accounts of the European Agricultural Guidance and Guarantee Fund (EAGGF) Guarantee Section. (5) The Conciliation Body first submitted a list of questions and then took evidence from the Commission's officers and the German authorities.

9 On 1 March 1996 and in the light of the differing positions of the Commission and the German authorities, the Conciliation Body arrived at provisional conclusions and proposed to both sides that they enter into a detailed dialogue with a view to reaching agreement on the facts.

10 By letter of 19 March 1996 addressed to the Conciliation Body, the German authorities stated their readiness in principle for fresh discussions with the Commission, but pointed out that discussions aimed at conciliation would be worthwhile only if the Commission was prepared to seek a constructive compromise.

11 By letter of 21 March 1996, the Commission's officers made known that they were maintaining their position and that, even though the Commission had no objection to such discussions, there was no real point in a further meeting.

12 In the circumstances, the Conciliation Body did not consider it worthwhile arranging such discussions. In its final report of 29 March 1996, it did not take a vote to decide in favour of one or other of the parties to the conciliation procedure and maintained its provisional conclusions.

13 It was against that background that the Federal Republic of Germany brought its action.

II - Preliminary remarks

14 The Federal Republic of Germany denies that the German control system fails to meet the requirements of Community law and challenges the accuracy of the findings of the Commission inspections. It considers that the flat-rate reduction is not therefore justified.

15 The Federal Republic of Germany also maintains that the Commission failed to take account of the arguments and evidence adduced both orally and in writing during the period that preceded the action.

16 According to the Commission, though it did not find the evidence furnished by the German Government convincing, it is not true to say that the Commission failed to consider it.

17 I take the view that even if, as the Commission maintains, it must be borne in mind that the flat-rate correction at issue relates to the inadequacies of the control system as a whole and not the inadequacies specific to the different stages in the intervention machinery, it is none the less incumbent on the Commission to establish the existence of each of the alleged inadequacies. Consequently, the main issue involved in this action, namely whether there are in fact deficiencies in the German control system, can be resolved only after detailed analysis of the evidence provided by the German Government in relation to the findings on which the Commission based its decision.

18 In that context, I consider it useful to cite immediately a number of points of case-law concerning the burden of proof. The Court of Justice has pointed out inter alia in a judgment of 1 October 1998 (6) that:

`only intervention undertaken in accordance with the Community rules within the framework of the common organisation of the agricultural markets is to be financed by the EAGGF (see Case C-48/91 Netherlands v Commission [1993] ECR I-5611, paragraph 14). In that respect, it is for the Commission to prove an infringement of the rules on the common organisation of the agricultural markets (see Case 347/85 United Kingdom v Commission [1988] ECR 1749, paragraph 16; Case C-281/89 Italy v Commission [1991] ECR I-347, paragraph 13; Case C-55/91 Italy v Commission [1993] ECR I-4813, paragraph 13; and Case C-48/91, cited above, paragraph 18). Accordingly, the Commission is obliged to give reasons for its decision finding an absence of, or defects in, inspection procedures operated by the Member State in question (Case C-8/88 Germany v Commission [1990] ECR I-2321, paragraph 23).

The Member State, for its part, cannot rebut the Commission's findings by mere assertions which are not substantiated by evidence of the existence of a reliable and operational supervisory system. If the Member State is not able to show that they are inaccurate, the Commission's findings can give rise to serious doubts as to the existence of an adequate and effective series of supervisory measures and inspection procedures (see, to that effect, Case C-8/88 Germany v Commission, cited above, paragraph 28).'

19 In support of its challenge, the Federal Republic of Germany also cites the failure of the Commission to charge to it any sum during previous financial years for which the German controls were carried out using the same system, even though the Commission had undertaken, or arranged for, detailed inspections.

20 The Commission explains that it refrained from charging sums to the Federal Republic of Germany for the financial years 1987 and 1988, even though the inspections carried out at the time had revealed shortcomings about which it had made recommendations to the German authorities. Examinations of the system carried out in 1991 and 1992 had, however, ascertained that those fundamental defects had not been remedied and further shortcomings had emerged. The Commission adds that the fact that it had apparently mentioned only certain imperfections in no way prevented it noting other deficiencies identified during subsequent inspections of the control system and taking them into account in imposing charges.

21 In regard to that argument, I share the Commission's view that the fact that it does not draw financial inferences from a finding of deficiencies pertaining to one financial year cannot deprive it of the right to draw such inferences in relation to subsequent financial years, particularly where those deficiencies have persisted. Deficiencies identified subsequently may also be taken into account in determining the level of the flat-rate correction, without there being any need for prior recommendations or warnings to have been issued in relation to those deficiencies.

22 Finally, the German Government claims that its position is borne out by the conclusions of the Conciliation Body.

23 The Commission challenges the German Government's interpretation of those conclusions and submits, more generally, that, pursuant to Article 1(2)(a) of Decision 94/442/EC, it is not bound by the conclusions of the Conciliation Body when adopting its final decision. The German Government does not take issue with that.

24 Alongside a lengthy description of the points on which the parties disagree, the conclusions of the Conciliation Body include the following passages:

`[T]he German control system operated during the period in issue, and in particular the use of BALM representatives, could probably have been improved, as suggested by certain comments made by the EAGGF during the 1987 and 1988 verification procedures, and as shown by the measures taken in the meantime for that purpose by the German authorities (see their letter No 714/1248/1992 of 6.10.1992). Those measures do not, however, allow of the conclusion that the measures applied previously were insufficient to avoid financial risk for the EAGGF.

As regards the question whether in fact the German control system meets the standard of effectiveness required to avoid such risks, the Conciliation Body has not yet been able to make a final assessment.' Indeed, in that connection, `the assessment of the facts pertaining to the inspection situation differs too greatly, according to whether it is the Commission officials or the German authorities that make the assessment.'

25 Admittedly, this provisional view of the Conciliation Body indicates that it in principle approves the control system applied by the Federal Republic of Germany. However, the Conciliation Body does acknowledge that a final evaluation can be made only on the basis of a closer analysis of the actual operation of the system. It is therefore for the Court to undertake that analysis.

26 The Conciliation Body subsequently raises the question of whether the German control system was sufficiently reliable to avoid risks should the slaughterhouse and the coldstore form part of the same group of companies. In that connection, the Conciliation Body states the following:

`[T]he Commission departments maintain that they had no evidence to substantiate the German authorities' assertion that the controls were carried out by BALM officials where the slaughterhouse and coldstore belonged to the same group of companies. The German authorities have transmitted to the Conciliation Body documents substantiating that assertion, on the basis of specific and limited examples. The Conciliation Body fails to understand why that vital evidence was not able to be entered into in greater detail and clarified once and for all during the earlier stages in the procedure. It can only invite the parties to clarify the issue as a matter of urgency.'

27 That extract from the conclusions shows that the Conciliation Body considered that the German authorities had demonstrated the accuracy of their assertion in certain specific cases, but leaves open the question whether a general conclusion can be drawn from those cases.

28 A third extract from the conclusions of the Conciliation Body includes a provisional assessment. After considering whether the incidents of fraud uncovered by the German authorities would have been avoided had BALM used its own inspectors exclusively, the Conciliation Body states that it is `inclined to the view that there are insufficient grounds for that assumption, given the criminal nature of the irregularities.' I shall return to that aspect of the problem below.

III - Controls on takeover (entry into storage)

A - Pleas in law and arguments of the parties

29 The arguments the applicant puts forward to challenge the Commission's summary report in this regard turn on five points.

30 The summary report indicates that, in accordance with Article 14 of Commission Regulation (EEC) No 859/89 of 29 March 1989 laying down detailed rules for the application of intervention measures in the beef and veal sector, (7) product takeover (that is to say transfer of ownership and the determining point for payment) occurs on the day the products enter the intervention centre. The Commission therefore criticised the frequent use of employees of the coldstores as BALM representatives for the purpose of weighing and certifying takeover of the goods. Furthermore, according to the Commission, even though BALM's instructions to its regional offices are specific as to the content of the verifications the inspectors have to make, they are silent as to the timing and frequency of controls to be conducted. The controls described could not reasonably be made by means of inspection after the event because the quarters of meat have already been weighed, frozen, packaged, placed on pallets and put into storage. The EAGGF therefore cast serious doubt on the reliability of a control system based on certificates issued by warehouse employees.

31 The summary report further indicates that, on this point, BALM initially emphasised the conflict of interest between slaughterhouses selling into intervention and coldstores designated to store the goods. However, the report makes clear that the EAGGF control showed that this could not be relied upon because slaughterhouses and coldstores sometimes belonged to the same group of companies. Therefore, while specifically acknowledging the positive elements in the German system, the EAGGF concluded that its main criticisms held good in relation to the absence of independent and comprehensive official controls by BALM at the point of takeover.

32 In order to justify the use of warehouse employees as representatives, the Federal Republic of Germany stresses that there are two phases of control at the point of takeover: the first stage takes place in the slaughterhouse and the second in the coldstore.

33 In the initial phase, in the slaughterhouse, the German Government stresses that the controls are carried out by BALM officials only and not by other persons delegated by BALM; during this phase, the BALM officials undertake comprehensive verification of the intervention requirements.

34 In the second phase, in the coldstore, the applicant maintains that the controls are carried out, as far as possible, by a BALM official. It does, however, acknowledge that BALM has also used representatives employed by the coldstore. In such cases, BALM would, however, always select experienced employees to act as its representatives and give them specific training. Tables and planks, placed in the coldstores, enable them to verify whether the quarters delivered may be taken over. If the representative is in any doubt, the BALM regional office immediately sends in an inspector who decides whether or not the quarters at issue may be taken into intervention. Furthermore, persons used as representatives are regularly monitored, supervised and trained by the BALM supervisor.

35 The Federal Republic of Germany points out that the controls carried out in the coldstore by the individuals appointed as BALM representatives took the same form and were on the same scale as those duly carried out by the BALM inspectors. Furthermore, the inspections carried out by the BALM representatives were subsequently verified by means of ex post inspections based on checks on documentation and physical inspections of the goods taken over, carried out by BALM inspectors, officials from the BALM regional offices and BALM supervisors. That procedure provided adequate security and effective protection against tampering. According to the German Government, the control system constitutes a whole and has to be viewed in its entirety. Assessing the second phase of controls in isolation would inevitably lead to the conclusion that deficiencies existed, if no account were taken of the inspections carried out at the slaughterhouse. The German Government further points out that in making its criticism, the Commission failed to take account of the two-tier system and considers that the Commission decision does not, consequently, provide an adequate statement of the reasons on that point.

36 The Commission's response is that it is not calling into question the controls at the slaughterhouse, which are neither the subject of the sums charged nor, consequently, of this dispute. Those controls cannot, however, compensate for the shortcomings in the German system identified as the stage of entry into storage. In the Commission view, no conclusions on compliance with the control requirements on entry into storage can be drawn from the regularity of the controls operated at the slaughterhouse. Nor can it be inferred from the regularity of the controls at the slaughterhouse that the controls at the coldstore should be less stringent. Those controls have their own crucial importance for the Community payments system. BALM itself recognised that importance by issuing special instructions concerning implementation of the controls.

37 In regard to the checks on documentation and the physical inspections carried out after the event, the Commission considers that while they are, admittedly, useful and necessary in relation to the BALM system of representatives, these ex post controls are of only limited scope. In point of fact, these are basically checks on written documentation which fail to take account of factors as important as, for instance, whether the quarters of meat delivered were properly cut up and dressed, whether the temperature required at delivery (plus or minus 7oC) was maintained and whether the quarters were properly weighed.

38 The Commission places particular stress on the risks bound up with the use of representatives, namely the impossibility of correcting at a later stage irregularities in the controls at the point of takeover and the conflicts of interest where the slaughterhouse and coldstore are in the same ownership.

39 In regard to the first of those risks, the Commission maintains that the controls required on entry into storage can only be carried out at the time the batches of beef are delivered to the coldstore. If, at that stage, the controls are not carried out or are not carried out properly, they cannot subsequently be corrected on the basis of checks on documentation or physical checks, even if carried out by BALM officials, because the batches of meat at issue have already been weighed, frozen and packaged, placed on palettes and put into storage. That highlights the special responsibility incumbent on the persons responsible for operating controls on entry into storage.

40 In regard to the risks bound up with conflicts of interest where the slaughterhouse selling the meat and the coldstore responsible for storing it form part of the same group of companies, the Commission is of the view that the German system does not pay proper attention to that conflict of interest. According to the Commission, it is quite obvious and perfectly normal that a conflict of interest should arise in the person of the BALM representative because that representative is both an employee of the coldstore and is responsible for verifying the regularity of the operations carried out by the coldstore when the meat enters into storage. If, despite that conflict of interest, the use of BALM representatives is unavoidable, the German authorities ought, according to the Commission, to have taken effective steps to ensure that the conflict of interest does not undermine the quality of the controls that have to be carried out at the time of entry into storage.

41 In that connection, the Federal Republic of Germany stresses that whenever, in the context of the intervention purchase of beef, the vendor was a linked undertaking (slaughterhouse/coldstore), BALM took special care to ensure that, as far as possible, the beef purchased from that undertaking was not put into storage in a coldstore linked to the vendor (slaughterhouse). In those rare instances in which, for logistical and/or organisational reasons, it had been necessary to depart from that principle, the controls on takeover at the coldstore had been carried out by a BALM inspector exclusively. In that connection, the Federal Republic of Germany cites six examples of linked undertakings and attaches, by way of an annex, the file on controls in relation to the undertakings Annuss (slaughterhouse) and Annuss (coldstore). It goes on to say that BALM had stipulated that this was the procedure to be followed (controls to be carried out by BALM officials exclusively) in the case of linked undertakings and had transmitted oral instructions to that effect to the regional offices.

42 The Commission casts doubt on the credibility of those claims by pointing out that, initially, the German authorities maintained that companies selling their products into intervention were prohibited by law from storing their own products. An inspection subsequently revealed that this claim did not reflect the true state of affairs and that linked undertakings were actually entering into such transactions with one another. According to the Commission, the German authorities than declared that, in the circumstances, the controls had to be tightened. Finally, in the Commission view, the Federal Republic of Germany is now claiming, in its application, that, in the case of linked undertakings, controls have always been carried out by BALM officials. Those are the reasons why the Commission considers those arguments unconvincing. Moreover, they are not mirrored in BALM's instructions to its regional offices. The instructions make no reference whatsoever to the special problem of linked undertakings. The Commission considers that production of the file on the controls carried out in relation to Annuss/Annuss alone cannot prove that the inspections were carried out systematically and entirely by BALM officials.

43 The Commission summary report indicates that doubts about the reliability of the German control system were augmented by the knowledge that in one irregular case revealed during the EAGGF audit, employee(s) of the receiving coldstore had removed certain cuts from the quarters between takeover and blast-freezing. In another significant case, classification stamps on beef had been fraudulently changed from `O' to `R' to allow 6.7 tonnes of otherwise ineligible goods to enter store. In yet another case, a false BALM stamp was used to allow fraudulent entry of 92 tonnes of beef to store. That case was not detected until after the goods were removed from store.

44 The Federal Republic of Germany contends that those three instances in which the rules were breached cannot cast doubt on the security and regularity of the control system. Although indicative of the scale of criminal intent, they were detected in full as a result of the controls carried out by BALM and other German authorities and those responsible were held to account, with the result that the Community suffered no financial harm.

45 The Commission responds that the fact that those three instances of harm were detected by BALM or other German authorities cannot be considered evidence of the reliability of the German control system either. It again points out that the 2% flat-rate correction is not based on those cases, but that, in its view, they clearly illustrate the shortcomings in the control system at the stages of takeover, stock management and the removal from storage of beef within the intervention system.

46 In that regard, the Commission maintains that the first case is a good illustration of the shortcomings in the German system. Had the controls been carried out regularly and in accordance with Community law at the time the goods were taken over, there would have been no possibility of pieces of meat being spirited away between delivery and freezing because those controls extended over the whole of that period, with the quarters of meat being subject to control from the time they were delivered until they were frozen. The fact that this fraudulent activity was subsequently detected does not alter the fact that the control on takeover was incomplete and inadequate.

47 In its defence, the Commission maintains that it has been established that the number of half-carcases taken over into storage frequently fails to tally with the number of half-carcases earlier classified at the slaughterhouse.

48 On that point, the Federal Republic of Germany responds that - as the German departments explained to the Commission in detail at the control stage - the reason for the discrepancies in the number of half-carcases quality-tested in the slaughterhouse is that, during transportation to the coldstore, faults (transport stains, for instance) appeared in the half-carcases destined for intervention, and these products were consequently rejected on inspection for entry into storage at the coldstore because the intervention requirements had not been met. Those discrepancies have nothing to do with deficiencies in the control system. Moreover, they are not relevant given that only meat taken into storage at the coldstore is covered by intervention.

49 In its rejoinder, the Commission takes the view that the explanations provided by the Federal Republic of Germany in this respect are perfectly plausible. However, according to the Commission, it is surprising that those anomalies were not mentioned in the reports on quality assessment and takeover of the beef, in the instances mentioned in the summary report in particular. Those cases also demonstrate that the existence of such irregularities did not necessarily result - as the German Government seeks to imply - in rejection of the goods in question.

50 As regards the general risks of tampering, resulting in good quality meat being replaced with meat of inferior quality, the Federal Republic of Germany claims that no complaints were received from purchasers in relation to the meat exported (95%). In its opinion, had such tampering occurred, the third party purchasers would have complained about it. In regard to the portion of stock destined for food aid (5%), there was no possibility of swapping the meat because it would be difficult to sell goods that had been cut up and packaged.

51 The Commission's response here is that there was a real possibility of tampering which became a virtual certainty in the case of linked undertakings, because of the possibility of introducing a false report into the system, without fear of subsequent discovery.

52 According to the explanations furnished by the Federal Republic of Germany, it is not possible to introduce a false takeover report into the payments system unless there is collusion between all the individuals and authorities involved in the takeover of the intervention beef and administrative control (slaughterhouse, BALM inspector/person appointed representative by BALM, sworn weigher, coldstore, BALM regional and central offices), because the requisite controls are always carried out. However, the Federal Republic of Germany points out that no system could provide adequate protection against extensive and large scale fraudulent activity of that kind, involving all of the individuals and authorities.

53 The Commission retorts that in its reply to a Commission letter of 13 April 1994 stressing the risk that false reports might be introduced, the German Government stated on 6 July 1994 that `in order to guarantee that the takeover reports of the BALM central office, its regional offices and the coldstore tally ... an internal control and adjustment procedure will be introduced parallel to the current control procedure - as suggested by the EAGGF.' According to the Commission, in making that declaration, the German Government acknowledged that the procedure operated hitherto did not fully protect the payments system against the introduction of false reports.

54 In its summary report, the Commission noted that `comparison of the statistics showed some stark contrasts in regional control activity. For example BALM Hamburg, with about 40% of national deliveries to store over 1991 and 1992, could demonstrate an official BALM presence during takeover for more than half of the deliveries in its region (38% in 1991 and 72% in 1992). The least important BALM regions of Berlin and Weimar showed a control rate of 80%. Against these statistics the regions of Mülheim and Mannheim which accounted for about 25% of national deliveries had controlled at a rate of only 20%. The second most important BALM region of Munich (about 28% of national total) had not conducted any controls at all in 1991.'

55 In its application, the Federal Republic of Germany claims that it had been demonstrated, for the 1992 financial year, that more than half (some 52%) of the 14 000 controls carried out in a coldstore were made by a BALM inspector and that, when carrying out his inspection, the BALM inspector filled out, in addition to the takeover report (part B), a form called the `Bericht über die Prufüng im Empfangsbetrieb' (report on control in the receiving undertaking) which he transmitted to BALM. The other controls (some 48%) were carried out, according to the Federal Republic of Germany, by BALM representatives, with the result that there was 100% control of takeover.

56 In the Commission's view, the statistics show that there were real shortcomings, even though the bulk of the controls were carried out by BALM officials themselves (52% overall for the 1992 financial year). The Commission does not dispute that in 52% of cases the controls were carried out by BALM officials, but draws attention to the regional discrepancies identified in the summary report which it considers to be significant in terms of the control activity of the BALM regional offices. In the case of Munich, the differences identified reveal the scale of the deficiencies in the German control system.

57 As regards the assertion concerning the BALM Munich regional office inspectors, in 1991, the Federal Republic of Germany states that it is false and maintains that it immediately set the record straight when the assertion - based on a misunderstanding - was first made during the conciliation procedure. According to the Federal Republic of Germany, the takeover reports prove that the Munich regional office had BALM inspectors carry out 16% of the controls on takeover made in the coldstores.

58 In that connection, the Commission accepts that the assertion contained in the defence implies that the BALM officials failed to carry out any inspections in Munich. However, according to the Commission, that is not the main issue. The Commission explains that it actually established, on the basis of regional control data, requested and transmitted by the German authorities, that Munich alone was not adhering to the normal control system and substantiated the verifications by BALM officials by transmitting the `control report in the receiving establishment.' Given that no plausible explanation was provided to justify the particular case of the BALM Munich regional office in Munich, the Commission had to infer that the implementation of the controls by BALM officials, referred to in relation to the Munich office, was not properly substantiated. According to the Commission, it has also to be pointed out in passing that, based on the statistics supplied by the German services, not 16% - as claimed - but only 11% of the inspections were carried out by the Munich regional office (which is responsible for 412 of the 3 747 takeover reports in total drawn up by BALM officials).

59 In the summary report, the Commission states that a close examination of a selection of control reports showed that they do not necessarily signify comprehensive control (i.e. from start to finish and incorporating all elements provided for in the national instructions) of a particular consignment. On the other hand, about 10% of the reports viewed by the EAGGF indicate some negative observation (eg bad cutting, too high a temperature etc) resulting in rejection (at the coldstore) of all or part of the consignment. That underlines the importance of official independent controls at the point of takeover, particularly where slaughterhouse and coldstore are effectively one and the same.

60 In its defence, the Commission explains that it is clear from consideration of the reports that they fail to meet the requirements of Community law. That, it claims, is quite apparent from an examination of the control reports transmitted - albeit belatedly - to the Commission. The reports fall into three distinct categories:

First category: reports on the controls on takeover made by the BALM officials themselves at the time the batches of beef are delivered: the reports frequently mention isolated deficiencies; that is fully in accord with current practice and what is to be expected, and suggests that the controls were exhaustive and punctilious.

Second category: reports on the ex post controls on takeover carried out by BALM officials. These reports confirm that the batches of beef in question were examined only after they had been frozen, and state specifically that, at that stage, it is no longer possible to undertake comprehensive controls of takeover.

Third category: reports on the takeover of the intervention goods made by BALM officials; these reports form part B of the `Report on the qualitative assessment and takeover of the beef' which is used, after the slaughterhouse findings have been recorded in part A, to confirm the takeover of the beef for storage. It should be noted that the BALM representatives do not enter any negative comments in part B of the report. That does not reflect what happens in practice and is, moreover, belied by the outcome of verification by the BALM officials. These are clear contradictions that confirm the Commission's finding of what it considers to be real shortcomings in the system.

61 The Federal Republic of Germany explains that this is an incorrect assessment of the documents. According to the Federal Republic of Germany, there are two documents: the `Report on the qualitative assessment and takeover of the beef', which the Commission describes as the control report, and the form entitled `Report on control in the receiving undertaking'. In the German system, it is only the control report which is authentic. The other is a BALM internal document.

62 Consequently, according to the Federal Republic of Germany, the Commission's claims are inaccurate in so far as they refer solely to the `Report on control in the receiving undertaking'. Furthermore, the Federal Republic of Germany considers it necessary to reject the Commission's claim that the takeover reports (part B) drawn up by the BALM representatives, and containing no negative comments, do not reflect what the Commission has found to be the case in practice. On the contrary, the Commission has produced only a very limited number of `Reports on control in the receiving undertaking' containing the findings of the BALM inspectors on the discrepancies that exist as compared with the normal control procedures. Those reports cannot therefore serve as proof of an alleged general lack of control consequent on the use of representatives.

63 In that connection, the Commission points out that, in its defence, it selected just a few examples to substantiate its arguments. The German Government allegation that it was possible to talk in terms of only a `very small number of reports' is therefore irrelevant. The Commission claims that, all in all, it found references to deficiencies in about 15% of the control reports analysed after the bilateral meeting of 20 January 1996, one of which noted `no BALM stamp'.

64 The Commission further points out that it is the German authorities themselves who put forward the `Reports on control in the receiving undertaking' to demonstrate that the German control system was working properly, after the Commission had ascertained that the takeover reports were frequently not drawn up by the German authorities, in accordance with their own instructions, and were sometimes not even signed. It is therefore logical that the Commission departments also should take into consideration the `Reports on control in the receiving undertaking'.

65 The German Government further points out that the form `Report on the qualitative assessment and takeover of the beef', used during the 1992 financial year for the takeover controls, did not include an appropriate section in which either the BALM inspectors or their representatives could record any unusual features that came to light during verification of takeover or minor deficiencies noted in relation to the beef to be taken over. Only the half-carcases duly taken over at the coldstore were entered in part B of the takeover report. Subsequently, and at the suggestion of the EAGGF, BALM recast the report to include a section for unusual features.

66 According to the Commission, that is a further illustration of the shortcomings of the German system at the time. It considers the absence of an appropriate section in the report to be a serious omission. And that omission is all the more serious if - as claimed by the German Government - the takeover report is in fact the main control document covering the stage of entry into storage. A section of that kind would have encouraged the person responsible for carrying out the controls to look carefully for any deficiencies in the meat delivered and, where appropriate, positively to confirm, by entering his signature, that the goods were of sound quality.

67 As regards the beef, in relation to which a number of `Reports on control in the receiving undertaking' reveal that it was inspected in the freezer tunnel, the Federal Republic of Germany claims that the meat was still fresh and not - as the Commission claims - frozen. The controls on takeover were able to be duly carried out because the process of freezing did not begin until the whole of the batch delivered had been put into the freezer tunnel.

68 According to the Commission, it is clear from two of the reports at issue in particular that the meat was not inspected until the day after its delivery or, indeed, the day after that. It is for the German Government to explain whether and how it is possible that, after a period of more than 24 hours, the fresh meat was still fresh and had not become frozen. Leaving that aside, the ex post controls do not, in any event, make it possible to check on whether the temperature of 7oC prescribed for delivery was maintained. Though the reports confirm that temperature, that confirmation is not based on controls by the BALM officials.

5. Is Regulation (EEC) No 2456/93 applicable?

69 The Federal Republic of Germany further maintains in its application that the Commission is basing its objection to the system of entry into storage operated in Germany - not specifically, it has to be said, but in substance - on Article 17 of Commission Regulation (EEC) No 2456/93 of 1 September 1993 laying down detailed rules for the application of Council Regulation (EEC) No 805/68 as regards the general and special intervention measures for beef. (8) In the German view, it is clear from the recitals of the regulation that the Commission itself accepted that the range of measures available to it in the sector was not sufficient to deal with the exceptional situation in 1992 (when intervention stocks of beef were at their highest since the establishment of the Community). However, the Federal Republic of Germany contends that this regulation did not enter into force until September 1993 and was not applicable to the 1992 financial year because Regulation No 859/89, then in force, did not contain an analogous provision. The Federal Republic of Germany objects that, using Article 8 of Regulation No 729/70 as a pretext, the Commission has applied retroactively to the 1992 financial year the more stringent requirements of Regulation No 2456/93.

70 According to the Commission, that objection is unfounded because the criticisms directed towards the German control system in respect of the 1992 financial year are based neither in fact nor in law on the rules put in place by Regulation No 2456/93. The Commission maintains that its criticisms are actually based on the fact that the Federal Republic of Germany failed to take the measures necessary to ensure that the transactions financed by the EAGGF were actually carried out and were executed correctly, in accordance with Article 8(1) of Regulation No 729/70.

B - Analysis

71 Are there deficiencies in the German control system at the point at which the intervention beef is taken into storage which might result in a minor risk of losses for the EAGGF?

72 Notwithstanding the arguments put forward by the applicant, I consider that question has to be answered in the affirmative.

73 It is clearly undeniable that, when controls are carried out at the slaughterhouse, the BALM officials make comprehensive checks on whether the intervention requirements have been met. Those checks cannot, however, compensate for possible deficiencies in the controls at the time the meat is taken over at the coldstore. It has to be noted that the regularity of the controls at the slaughterhouse is not relevant to an assessment of the regularity of the controls at the time of takeover at the coldstore. In accordance with Article 14 of the abovementioned Regulation No 859/89, the critical point, particularly in relation to the transfer of ownership of the intervention goods and payment, is the point of takeover. The Commission is therefore right to state that the controls at the coldstore have their own crucial importance for the Community payments system.

74 As regards those controls, it is significant that, as the German Government has itself acknowledged, were they assessed in isolation, the inevitable conclusion would be that deficiencies existed. (9)

75 In that connection, the Commission's findings based on the control reports reveal that the representatives were not sufficiently searching in the controls they carried out. It is in fact apparent that where it was the BALM inspectors who carried out the controls, the reports frequently mention isolated deficiencies, whereas the reports by the BALM representatives contain no negative comments. That inconsistency is borne out by later controls, carried out by the BALM officials, 15% of which refer to shortcomings involving the controls carried out by the representatives.

76 Germany's argument, according to which the ex post control reports drawn up by the BALM officials are simply a BALM internal document, and only the control report made at the point of takeover is authentic, cannot be accepted. In fact, as the Commission points out, the German authorities themselves put forward those documents to show that their control system was working properly. Consequently, if the documents provide arguments that strengthen the Commission's case, the Commission is entitled to rely on them.

77 Furthermore, I share the Commission's view that the requisite controls can be carried out only at the moment when the batches of beef are delivered to the coldstore, since the controls carried out subsequently by BALM officials cannot make good possible shortcomings in the controls made at the time of takeover. If the BALM inspector is not present at the time of takeover, he will in fact inspect the meat either when it is in the freezer tunnel (Schockraum) or when it has already entered storage.

78 If the meat is inspected in the freezer tunnel, it is hard to imagine that, after 24 hours has elapsed, the meat is not frozen to such a degree that it is no longer possible to verify its quality with sufficient accuracy, to say nothing of its temperature on arrival.

79 If the inspection is made after the meat has entered into storage, detailed checks on its state on entry to the coldstore are still more impossible because the meat has been completely frozen, packaged and placed on a pallet in the coldstore.

80 These subsequent documentary and physical checks are therefore of limited use in establishing the regularity of the transactions.

81 Furthermore, if it is thus established that the main controls are those carried out when the meat is taken over, the fact that the report used for takeover controls in 1992 did not contain an appropriate section in which to record any unusual features that emerged on verification of takeover or minor deficiencies noticed in the meat to be taken over, constitutes a further defect in the German system at the time. That would be all the more the case if that document were in fact the only authentic document, as the German Government claims. Moreover, the latter recognised that defect and remedied it by introducing an appropriate section for subsequent financial years.

82 More difficult is the question of the conclusions to be drawn from the fraud detected. As I mentioned at the outset, the Conciliation Body is `inclined to the view' that because of the criminal nature of the irregularities, they could not have been avoided even if BALM had used only its own inspectors. That may be so. On the other hand, it seems to me, that the fraudulent substitution of category `R' stamps for category `O' stamps and the use of a forged BALM stamp to allow 92 tonnes of meat into intervention would be much more difficult if the BALM officials were in attendance when the goods were delivered to the coldstore. As the German Government points out, these cases of fraud cannot, of course, call into question the security and regularity of the German control system as a whole. But they do reveal the vulnerability of the system resulting from the use of representatives.

83 That vulnerability becomes all the more apparent if one takes into consideration the relatively limited number of controls carried out by some BALM regional offices at the time of takeover. Even though the total percentage of controls carried out directly by BALM officials was 52%, there is reason to fear that the low level of official and independent controls carried out by the Mülheim, Mannheim and, above all, Munich offices, made it easier for the unscrupulous to engage in fraud.

84 However, as regards the risks of fraud in the case of linked undertakings, I am surprised at the Commission's assertion that only in its application did the German Government maintain that controls are always carried out by BALM officials in such cases. In point of fact, the summary report itself makes clear reference to the oral instructions given to the regional offices, according to which takeover has exclusively to be checked by a BALM official in cases in which the slaughterhouse and coldstore belong to the same group of companies.

85 However, the fact that these were purely oral instructions - which the Federal Republic of Germany does not dispute - is a ground for criticism. Where a coldstore and slaughterhouse are linked, it is crucially important for the regularity of takeover operations that these controls be carried out. In issuing its oral instructions, BALM recognised the importance of this but failed to give them the force of written instructions. It can therefore be considered that, because BALM's instructions were oral, they were not as effective as they could have been and that fact may therefore also be taken into account in assessing the defects of the German control system.

86 As regards the possibility of introducing a false report into the system if the two undertakings are linked, it has to be said that the possibility exists. The risk that that kind of fraudulent manoeuvre will not be detected is, admittedly, a minor one, but it nonetheless exists. The German authorities themselves acknowledged that to be the case by announcing that they were going to introduce a procedure designed to guarantee that the copies of the takeover reports in the possession of the BALM central office, its regional offices and the coldstore tallied in every detail.

87 Finally, as regards the German Government's argument criticising the Commission for retroactively applying Article 17 of the abovementioned Regulation No 2456/93, I can only endorse the Commission view. That criticism is justified neither in fact nor in law. It is true that Article 17 sets out for the first time, in detail, the procedure for takeover. But even if the Community legislation did not yet contain provisions as specific as this in this area, the fact remains that Article 8 of the abovementioned Regulation No 729/70 itself required the Member States to take the necessary measures to:

- satisfy themselves that transactions financed by the Fund are actually carried out and are executed correctly;

- prevent and deal with irregularities.

88 That provision, which constitutes, within the field of agriculture, an expression of the obligations imposed on the Member States by Article 5 of the EC Treaty, sets out the principles according to which the Community and the Member States are to organise implementation of the Community decisions on agricultural intervention financed by the EAGGF, as well as the fight against fraud and irregularities in relation to those operations. It imposes on the Member States the general obligation to take the measures necessary to satisfy themselves that the transactions financed by the EAGGF are actually carried out and are executed correctly even if the specific Community act does not expressly provide for the adoption of particular supervisory measures. (10)

89 It is in the light of that obligation, therefore, that the German authorities were supposed to ensure the effectiveness of the controls at the time the beef was taken over.

90 The above analysis has shown to be unconvincing the arguments the German Government has used to try to prove that the shortcomings on which the Commission based in particular the flat-rate correction at issue did not exist.

91 Finally, as regards the complaint that the statement of the reasons was inadequate, it has only to be borne in mind that decisions on the clearance of accounts do not require detailed reasons, if they are taken on the basis of summary reports as well as correspondence between the Member State and the Commission, which implies that the government concerned has been closely involved in the process by which the decision came about and is therefore aware of the reason for which the Commission considered that it must not charge the sums in dispute to the EAGGF. (11)

IV - Controls during storage

A - Pleas in law and arguments of the parties

92 The applicant's arguments on this issue turn on four points.

93 Article 3 of the abovementioned Regulation No 618/90 reads as follows:

`The accounting inventory shall be verified by the storekeeper during the last two months of the financial year.

This verification shall entail verification of the physical presence of the goods in accordance with a form, a model for which is given in Annex II.

The verification shall be carried out in the presence of an official of the intervention agency.

If, however, the official of the intervention agency is not present, he shall carry out the verification on his next visit, taking account of stock movements occurring in the interval.'

94 The form, a model of which is given in Annex II of the regulation, contains four columns with the following headings: `weight entered', `weight checked', `weight verified' and `observations'. The following appear at the bottom of the form: `weight checked: establishment of physical presence of goods (Article 3)' and `weight verified: batches selected and weighed (Article 4).'

95 In its summary report, the Commission noted that, in the inventories, there were never any differences between the book weight and the checked weight. In the coldstores visited it was confirmed that stocks had not been verified, as required by Article 3, but had simply been transposed from the stock accounts of the coldstore.

96 The Federal Republic of Germany contends that Article 3 of the abovementioned Regulation No 618/90 does not require that the weight be verified as part of the annual inventory control because, pursuant to Article 4 of that same regulation, physical verification takes place subsequently. Consequently, according to the Federal Republic of Germany, contrary to what the Commission alleges, no difference in weight is able to be established in the annual inventory declarations at the time the physical presence of the goods is verified.

97 The Commission maintains that the disappearance of any pallet or piece inevitably results in a weight loss which must be entered in the inventory report. The fact that no such entry exists confirms what employees of the coldstore, whom the Commission met with during on-the-spot verification, told the Commission official directly, namely that the storekeeper did not verify the physical presence of the goods but directly transposed the data on quantity and weight entered in the coldstore stock account. That, according to the Commission, is the real flaw in the inventory system because the annual inventories drawn up by the coldstores never indicated any discrepancy between the quantities and weights recorded in the accounts and the quantities and weights checked.

98 The Federal Republic of Germany contends that it is for the Commission to prove those allegations and puts forward, by way of proof, the evidence of some of the BALM regional office inspectors. The Federal Republic of Germany then explains that, every year, the storekeeper is given written instructions to comply with the requirements relating to the drawing-up of the annual inventory. By signing the declaration on the annual inventory for the intervention goods, the storekeeper certifies and establishes that the control has been properly carried out. The regulation does not provide for additional proof or documentation confirming that Article 3 was complied with when the accounting inventory was checked.

99 In that connection, the Commission replies that the mere obligation to verify the physical presence of the goods in itself results in the need for documents that verifiably establish that the controls have been carried out. That is why the proof required by the Commission is not `additional' proof but simply proof that verification has taken place.

100 According to the Federal Republic of Germany, at the point at which the physical presence of the goods is ascertained, there is no need physically to check the weight, by weighing the goods as part of the inspection provided for in Article 3; all that has to be done - in accordance with the definition of the phrase `weight checked' - is to record the weight which was verified on the spot as part of the process of establishing the physical presence of the goods pursuant to Article 3. Given that there is no further control of weight, by means of a subsequent weight check, the weight that appears in the columns on the form relating to the weight entered and the weight checked are always going to tally if on-the-spot stock verification has not identified any mistake in the inventory.

101 On that point, the Commission maintains that, contrary to the arguments it is now advancing, the Federal Republic of Germany has not ruled out the possibility that a practice of transcribing exists in the coldstores and, in response to a comment from the Commission stressing that risk, the Federal Republic of Germany stated that it intended `checking the storekeepers' compliance with the provisions more closely in the future.' Moreover, the Commission points out that, since the figures were simply transcribed, it is perfectly natural that no quantitative inaccuracy could be identified but, in its view, this cannot be used as a basis for concluding, as the German Government would have it, that the controls were regular.

102 Article 4 of Regulation No 618/90 provides:

`[O]n the basis of the accounting inventories, the official of the intervention agency shall check on the spot:

- the procedure used by the storekeeper to draw up the inventories

- the accounting data of the warehouse (entries, removals, balance of stocks) and the outcome of the storekeeper's annual inventory.

The official shall compare the abovementioned data with the quantitative data in his own accounts and with the monthly accounting inventories.

Where there is agreement between the accounts, he shall make a physical inspection of at least 5% of the quantities stored in accordance with the methods set out in Annex III for each product.

Where accounts differ or where the physical inspection reveals a discrepancy not clearly justified, an additional percentage of the quantities in intervention storage must be checked in accordance with the same method. This check must be continued by adding an additional 5% each time to the quantities involved until the differences or discrepancies are explained.

The intervention agency must be represented by an auditor or controller approved by a Member State.

The inspection shall be carried out by a person independent of the storekeeper.'

103 Annex III, section VI, of the same regulation lays down the following methods for the physical inspection of beef with a view to drawing up an inventory:

`1. Selection of batches representing 5% of the total quantity in public intervention storage. The batches to be checked may be selected prior to the inspection of the warehouse on the basis of the accounting data of the intervention agency, but the storekeeper may not be informed.

- for bone-in meat:

- identification of the batches and verification of the number of pieces,

- verification, for each type and/or quality, of the weight of 20% of the pieces,

- visual check of the condition of the packaging;

- for boned or boneless meat:

- identification of the batches and verification of the number of boxes,

- weighing of 10% of the pallets or containers,

- verification, for each cut, of 10% of the boxes from each pallet or container,

- visual check of the contents of the boxes and the condition of the packaging in each box.

The pallets/containers shall be chosen following the different cuts stored.

104 In its summary report the Commission notes that it was not able to establish that a physical inspection of at least 5% of the meat in store had taken place, in accordance with Article 4 and Annex III to the regulation. The report describes the difficulties the EAGGF encountered in undertaking the controls of the inventories: the storage facilities were tightly packed, there were problems in ascertaining the actual quantities visually inspected and in determining the number of pieces on each pallet, there were no controls designed to guarantee that the selected pallets were those weighed, and so on.

105 In its application, the Federal Republic of Germany disputes the Commission's assertion that the controls provided for in Article 4 and Annex III were not duly carried out. It maintains that a certain number of batches, corresponding to at least 5% of the total quantity stored, were selected and that the composition of those batches was verified on the spot. It adds that, in the course of the additional verification covering 20% of the quantity (5%) already checked, weight was checked by the BALM official who selected on the spot the 20% of pieces to be inspected (selecting the pallets). The signatures of the BALM official and the storekeeper on the annual inventory report - which is accompanied by the relevant stock report that lists separately the batches stored that have been inspected and physically weighed - certify and establish the outcome of the physical inspection of the meat stored. According to the German Government, further requirements cannot be justified under Community law.

106 The Commission does not dispute that the weight of the 20% of the 5% sample already inspected was determined in accordance with Community law, but points out that this percentage accounts for only 1% of the total quantities stored. According to the Commission, that element of the controls was not criticised, which is why the observations on that point contained in the German Government application are irrelevant. According to the Commission, it is in relation to the physical inspection of at least 5% of the meat stored that there are serious deficiencies in the inventory control system. In that connection, the Commission maintains that the claims by the Federal Republic of Germany do not square with the actual circumstances encountered by the Commission officials when carrying out their own inspections. In point of fact, according to the Commission, the data on weight simply makes it possible to establish from which batch or category of goods the 1% of total stock was selected for weight control. The Commission points out, that in checking on the inventory controls carried out by the BALM officials, its own inspectors ascertained that it was impossible accurately to assess which quantities had been subject to visual checks. According to the Commission, the inventory reports did not make it possible to verify which and how many pallets had been weighed; the number of pieces per pallet was not counted; no checks were carried out to establish whether the pallets selected for weight controls were the pallets actually weighed; the pallets weighed by the BALM officials were impossible to identify; and no piece was unwrapped to compare the frozen meat or stamps with the data on classification and category. Moreover, according to the Commission, the `discrepancies' column in the inventory reports had frequently been left blank or not been filled in properly. In the Commission view, it is not possible to determine from the inventory reports the exact quantity that has been subject to physical inspection.

107 When the Commission asked BALM for documentary evidence it was unable to produce its inspectors' working plans showing the precise number of pallets and pieces inspected with corresponding weights. The Commission maintains that when, in relation to the use of Community resources, Community law lays down specific controls, such as the physical inspection of 5% of the beef stored, the competent national authorities must ensure that it is possible to verify whether, in the context of the procedure for the clearance of accounts, and on the basis of adequate documentation, that Community requirement has been met. Since that has not happened in this case, the burden of proof falls to the German Government and, according to the Commission, it is not sufficient here to rely on declarations by BALM officials or inspectors.

108 The Federal Republic of Germany claims that the actual quantity that has been physically inspected in accordance with Article 4 of Regulation No 618/90 appears in every annual inventory report, in the fourth column of the annex entitled `Stock report', with an exact record of weight.

110 According to the Commission, the German Government is describing a procedure which in theory makes it possible to identify the pallets inspected but whose actual implementation is in no way guaranteed. According to the Commission, its officials noted, for instance, in the course of their inspections, that the BALM inspectors often give the pallets selected a special number (from 1 to 20 and so on). In the Commission view, that approach made any further inspection impossible. In the Commission view, the use of pallet cards is certainly in principle a useful means of identifying the pallets, but in order for it to be possible to carry out controls using those cards, the beef would have actually to be still on the pallet. The Commission points out that if this is not the case, it is because clear instructions have not been provided.

111 In its summary report, the Commission states that

`national instructions regarding the execution of the 5% physical controls are not explicit but simply refer to the annex of the governing Community regulation.'

112 The German Government considers that the Commission is mistaken here because BALM drew up administrative instructions on the control of the annual inventory of agricultural products in public storage for the purpose of implementing Regulation No 618/90. In the field at issue here, it is sufficient that the administrative instructions refer to the provisions of Annex III, section VI of the abovementioned Regulation No 618/90, because that annex describes all the stages of control needed for the physical inspections to be properly executed at the time of the annual inventory and is perfectly comprehensible to the BALM officials, as it stands.

113 In the Commission view that is sufficient neither to guarantee that the physical inspection required under Community law is executed correctly nor to prove that it has been executed correctly, if necessary.

114 In its defence, the Commission notes that the inspections carried out in two coldstores provide a good example of the anomalies and deficiencies identified in practice. In the first coldstore, serious deficiencies were established in regard to three pallets: whereas the pallet card listed 12 quarters of meat with a gross weight of 1 129 kg, the inspection found only 11 quarters weighing 1 021 kg in total on the pallet. An unattached label, most probably belonging to the missing twelfth quarter, was lying on top of the 11 quarters. According to the card of another pallet, it contained 11 quarters with a total weight of 1 104 kg, subsequently amended to 1 094 kg. An unnumbered pallet contained three quarters with no indication of weight. Furthermore, in the second coldstore, one of the visible pallets held a quarter from which a large metal hook, that had probably been forgotten and included in the gross weight recorded at the point of entry into storage, was protruding.

115 The German Government explains those incidents as follows: in the first coldstore, the twelfth quarter was not missing but had been moved to another pallet because putting 12 quarters on one pallet was considered an overload. Because of an oversight, the pallet card was not amended. In regard to the second pallet mentioned by the Commission, the change to the weight merely corrected a mistake made when the original card was issued. As regards the three quarters for which no weight was indicated, these were goods that had not been able to be loaded on that day. Even if there had been a failure to mark the pallet, the place in which it was stored clearly linked it to the delivery from which it came. As regards the metal hook, the German Government notes that this was the first such incident despite the hundreds of thousands of quarters processed. In that particular case, the hook was removed and its weight deducted.

116 The Commission also points out that, pursuant to Article 26(1) of the abovementioned Regulation No 859/89, the meat is to be placed and kept in store in easily identifiable lots. In the coldstores visited, that requirement had been met in part only through the use of pallet cards; it was in fact noted that the premises were tightly packed, visual inspection was difficult because of the lack of access aisles and there were apparently no storage plans to show where goods were kept. Consequently, only the pallets at the front of the warehouse could be properly, that is to say rapidly and easily, inspected.

117 As regards the tightly packed nature of the coldstores, the Federal Republic of Germany considers that, although the warehouses were to some extent tightly packed during the 1992 financial year - because of the very high level of purchases - the stocks were properly inspected. In those warehouses that were very tightly packed, pallets were removed, according to the Federal Republic of Germany, to allow pallets from the rear of the warehouse to be inspected. According to the Federal Republic of Germany, the Commission's complaint on this point fails to take any account of the exceptional circumstances of 1992.

118 Finally, as regards the impossibility of identifying the pallets, the Federal Republic of Germany claims that the Commission is basing this on an isolated incident in which the BALM inspector responsible failed to monitor the removal from store of some of the pallets he had selected in the warehouse. However, according to the Federal Republic of Germany, no general conclusion can be drawn from this exceptional case and nor can it be seen as proof of an alleged lack of control characterising all the inspections carried out during the 1992 financial year, pursuant to Article 4 of the regulation.

119 The Commission maintains that two factors contradict that argument:

- The irregularities identified in the first case were so obvious that the EAGGF was able to uncover them simply on the basis of random sampling. More serious still, on the day on which the EAGGF made its inspection based on sampling, BALM employees had earlier carried out an inventory control during which those obvious irregularities were not picked up.

- In addition, the German Government was also unable to produce in its reply working papers drawn up by the BALM officials which, had the inventory controls been properly carried out, would have informed the Commission of the precise number of pallets checked and the number of quarters of beef stored on them. The Commission points out that it has always said that, notwithstanding deficiencies identified in specific cases, it would consider such working papers adequate proof that inventory controls that were regular overall had been carried out. However, the Commission has yet to be provided with any such working paper.

B - Analysis

120 As regards, firstly, the point at issue concerning entry in the inventory, I agree with the German Government that Article 3 of the abovementioned Regulation No 618/90 does not require that the meat be weighed. The Commission has also acknowledged that. The dispute therefore simply concerns whether, as the Commission maintains, when the annual inventory was drawn up, the storekeepers did no more than transcribe stock counts or whether, as required by the abovementioned Article 3, they actually verified the physical presence of the goods.

121 The Commission is, of course, right to point out that the loss of a pallet or piece will inevitably result in a difference between what is recorded in the inventory report and the stock counts. In this case, however, it is common ground that, for the 1992 financial year, no quantity was identified as missing when the inventory was drawn up. Therefore, the fact the Commission finds surprising, namely that the annual inventories showed no disagreement between the quantities and weights recorded in the accounts and the quantities and weights checked, seems to me perfectly plausible.

122 It remains to be ascertained whether, as required under Community law, the warehouse employees verified the physical presence of the goods. In that connection, the Commission bases its argument that this verification was not systematically carried out on employee statements. To rebut that, the German Government puts forward the evidence of BALM officials. However, the officials were in a position to provide that proof only in those cases in which they were actually present when the inventory was drawn up. In the other instances, it is not possible to challenge the evidence cited by the Commission according to which there were cases in which the storekeepers merely transcribed the figures in the stock accounts. Moreover, in a letter of 6 July 1994 addressed to the Commission's Director General for Agriculture, the Federal Minister of Finance acknowledged the following:

`... it has to be said that the BALM inspectors are not present in every case when the inventory is being drawn up. In accordance with Article 3(3) of the abovementioned regulation, the presence of representatives of the intervention agency is not, moreover, required. That is why it is not possible to comment on whether the storekeeper verifies the actual presence (of the goods) or simply copies (transcribes) the quantity entered in the stock accounts. The storekeepers are, however, given an annual reminder of the provisions on the drawing-up of inventories under Regulation (EEC) No 618/90. In future, BALM will monitor storekeepers' compliance with the provisions more closely.'

123 As regards, secondly, the inspection of the inventories by the BALM representatives themselves, the Commission's main criticism of the German authorities is that the annual inventory reports do not provide proof of compliance with Article 4 of Regulation No 618/90 as regards the obligation physically to inspect 5% of the quantities stored. Those documents ought to show that the physical inspection procedure laid down in Annex III of the regulation has been followed step by step and in every detail.

124 However, the annual inventory reports, including the stock reports that accompany them, do not provide that information. As the Commission rightly maintains, the weights recorded in the fourth column of these reports merely indicate from which batch or category of goods the 1% of total stock was selected for weight control.

125 Like the Commission, I take the view that where, in relation to the use of Community resources, Community law requires specific controls, as in this case the physical inspection of 5% of the beef in storage, the competent national authorities have to ensure that it is possible to verify compliance with that requirement in the context of the procedure for the clearance of accounts, on the basis of adequate documentation or references. It is therefore the lack of such documentation or references that constitutes the deficiency. In its letter of 6 October 1995, the Federal Ministry for Food, Agriculture and Forests stated, moreover, that in future `the controls to be carried out during storage will, as you suggest, be fully documented.' The ministry thus acknowledged that a gap existed here.

126 As regards, thirdly, the Commission's complaint concerning the absence of national instructions regarding the procedure to be followed by inspectors when carrying out the physical inspection, that procedure, I note, emerges clearly from Annex III to the regulation. Since that annex was attached to the administrative instructions, I consider that the inspectors were quite definitely in a position to understand what was required of them. It has, however, to be noted that the German authorities ought to have included in their instructions an obligation to draw up detailed reports on the physical inspection of 5% of the meat stored.

V - Controls on removal from storage

A - Pleas in law and arguments of the parties

127 Commission Regulation (EEC) No 147/91 of 22 January 1991 defining and fixing the tolerances for quantity losses of agricultural products in public intervention storage, provides that the tolerance limit for quantity losses resulting from normal storage operations carried out in accordance with the accepted rules is fixed as a percentage of the actual weight, without packing, of the quantities entering storage (Article 1(1) and (2)). According to Article 2 of the regulation, that percentage is set at 0.6% for beef.

128 In the case of non-identified losses of beef, in excess of the tolerance limit of 0.6% fixed by Regulation No 147/91, the Member States have to credit to the EAGGF a sum substantially higher than the sales value of the quantities lost.

129 Furthermore, unboned beef coming from public storage is to be sold, pursuant to Commission Regulation (EEC) No 2173/79 of 4 October 1979 on detailed rules of application for the disposal of beef bought in by intervention offices and repealing Regulation (EEC) No 216/69, according to the gross weight actually ascertained.

130 According to Article 18(2) of that regulation, `delivery shall be taken of the goods in accordance with the intervention agency rules on release from storage ...'

131 The Federal Republic of Germany gives clear expression to the above provision of Regulation No 2173/79 in paragraph 13.5.1 of Bekanntmachung Nr. 55/87/31 über die Allgemeinen Bedingungen für den Verkauf von Rindfleisch aus Interventionen of 8 September 1987 (Notice No 55/87/31 on the general conditions of sale for beef from intervention) which provides that, in the case of frozen beef, the delivery weight of any category of goods is to be established, on delivery, by a qualified weigher using accurately calibrated scales.

132 According to the second paragraph of section 13.5.1 of Notice No 55/87/31: `weighing records shall be used to indicate gross weight. The total weight entered in the weighing record shall be rounded up or down to the nearest kilogram in the case of quarters of meat and to the nearest one hundred grams for cuts. The weigher shall sign and stamp the records.'

133 According to the summary report: `Any losses of quarters are considered to be over and above the tolerance and lead to automatic credits to (the) EAGGF. It is clear therefore that Member States have an obligation to ensure that beef removed from intervention stock is correctly counted, weighed and reported for accounting and payment purposes.'

134 The Commission also notes the following in the summary report: `- The EAGGF found no evidence of BALM presence at the time of removal operations because neither the daily stock movement reports (Tagesmeldungen) nor the weighing records bear the signature of BALM inspectors. BALM has stated that it sometimes uses so-called BALM representatives to control removals because for staffing reasons BALM inspectors cannot be present at every removal. - The EAGGF established from its file control that the coldstore and purchaser often belong to the same company. In one case noted from its file audit, the EAGGF initially concluded that there had been no weighing of a particular batch of beef removed from a particular coldstore. From documents viewed it appeared that precisely (i.e. to the kg) the same quantity had been removed as had entered the previous year. While insisting that there was a 100% weighing of all beef leaving intervention, the German authorities have agreed that there was some misbooking of two different beef categories incorrectly accounted for together and so there was no weight loss in this case. Other cases noted by the EAGGF where there was not a kilogram of difference between entry and removal have been similarly explained. In the absence of weighing procedures duly controlled by BALM, the EAGGF is not convinced of the arguments and evidence put forward. It is clear that a highly significant proportion of intervention beef was removed from store without the presence of BALM.'

135 According to the German Government, the requirements of Community law concerning the removal of the beef were respected. Employees of the coldstore, who had been selected, trained and appointed by BALM, were also used, alongside BALM inspectors, to carry out the controls on removal of the beef. Even in those cases in which representatives were used, proper removal of the beef was guaranteed. Moreover, the purchasers would ensure that the meat auctioned and paid for was properly delivered in terms of both the quality and quantity of the beef because BALM's conditions of sale do not allow complaints to be made later. In those cases in which removal took place in the coldstores belonging to the same group as the purchaser, verification was made by BALM inspectors only.

136 According to the Commission, it is true that there is no legal requirement that controls on removal from store should be carried out by employees of the intervention offices only. There is, however, a general legal obligation requiring that the Member States ensure the accuracy of the quantities of meat removed from intervention stocks, given that the sums to be charged to the EAGGF are based on those quantities. In the Commission view, if the controls at issue are not carried out by BALM officials, the control system generally may be the subject of manipulation and abuse.

137 Even if - as the German Government maintains - the storage contracts entered into with the coldstores lay down control requirements, that such a requirement exists does not, in the Commission's view, mean that it is always complied with to the extent necessary. Furthermore, the German Government has still to prove that the conditions contained in the storage contract met Community control requirements in full. Moreover, as regards the subsequent controls carried out by BALM, the Commission considers that these merely involve inspection of documents and are insufficient to replace the on-the-spot controls required. As regards the particular problem of the linked undertakings, the Commission would refer back to the pleas and arguments it made in relation to the controls on entry into storage.

138 As regards the risk of manipulation, the Federal Republic of Germany maintains that such a risk is not conceivable because there is no Community market for frozen quarters of young calves, with the result that, were they placed on the market, questions would be raised as to the meat's origin. Furthermore, it would not make economic sense to replace that kind of meat with beef. In point of fact, the price of frozen quarters of young calves is in no way higher than that of quarters of fresh beef; the process of freezing causes a depreciation in the value of veal which is always greater than the price differential between the cost of frozen veal and that of beef. In addition, it is easy to distinguish beef from veal. Any purchaser of intervention meat would be able to tell the difference. Moreover, the purchasers of BALM stocks have made no such complaint.

139 As regards the exhaustive description of the controls on removal from store that the German Government provides in its rejoinder, the Commission maintains that this is largely confined to general observations on the `normal' progress of the control procedure but fails to address the specific deficiencies on which the Commission bases its findings.

140 The Commission draws attention to the findings of the summary report, according to which, in several instances, the EAGGF established that there was not a kilogram's difference between the weight recorded on removal from store and the weight of the goods that had entered into store the previous year, which is unusual in the case of goods stored frozen, where weight loss frequently occurs. According to the Commission, it has therefore to be presumed that the weight losses or increases were not properly declared because of an absence of controls.

141 According to the Federal Republic of Germany, once BALM had looked into the cases cited by the Commission, in which exceptionally, in its view, no difference in weight appeared between entry into and removal from storage, there were only two instances in which it was established that weight on removal from storage was the same as weight on entry into storage. Those particular cases were explained to the Commission. According to the Federal Republic of Germany, the explanation was that this anomaly was the result not of manipulation when the weight was recorded or a failure to weigh the goods but of the fact that two different categories of beef and veal were accounted for together by mistake, so that there was in fact no weight loss. According to the Federal Government, it is perfectly possible that, in specific cases, weight on entry into storage is exactly the same as weight on removal from storage.

142 According to the Commission, this attempt at an explanation is unconvincing and has not, moreover, been substantiated. The Commission therefore continues to consider the cases that have arisen to be clear evidence of shortcomings in controls on removal from storage in Germany, though these specific cases are not in themselves the basis for the 2% flat-rate correction of expenditure.

B - Analysis

143 As regards the controls on removal from store, the factors set out below seem to me decisive in reaching a determination.

144 I would first draw attention to the undisputed fact that neither the Community legislation applicable in 1992 nor the legislation that replaced it require that employees of the national intervention agency be present when goods are removed from store. There can therefore be no question of automatically concluding from the total or partial absence of such employees a deficiency in the control system capable of justifying a refusal to charge 2% of expenditure to the EAGGF.

145 However, one can only endorse the Commission's view that BALM employees must always be present in cases where the coldstore and purchaser of the meat belong to the same group of companies. Furthermore, that principle has not been challenged by the Federal Republic of Germany. The latter has merely objected that, in such situations, a BALM official was always in attendance. For its part the Conciliation Body concluded that the documents transmitted to it by the German authorities confirmed that claim `on the basis of specific and limited examples.'

146 In the light of the categorical assertions of the Federal Republic of Germany and the fact the Commission departments did not exhaustively verify all cases of this nature, it can reasonably be supposed that BALM officials were present when meat was delivered to undertakings linked to the company that owned the coldstore.

147 However, as I noted above in regard to the takeover of the meat by the coldstores, the German position is undermined by the fact that the written instructions given to the inspectors do not make their presence a formal requirement. In that sense, the control system is to an extent deficient.

148 In regard to all the other cases, that is to say those in which removal from store takes place under the supervision of BALM representatives (in the absence therefore of BALM officials), the Commission has identified cases in which the weight of the goods delivered was exactly the same as that of the goods taken over. An example of this appears in a report by EAGGF inspectors, transmitted by the Commission to the German authorities by letter of 13 April 1994 (No VI/014852, annex 3 of the application), page 10 of which states that 273 pieces of meat were weighed on entry at 26 591 kg and weighed exactly the same on removal. It is, however, clear from the legislation that, when they are entering storage, the pieces of meat have to be weighed before being packaged, whereas on removal from storage their gross weight has to be recorded, that is to say including the packaging. Now it is hardly likely that packaging will make up exactly for the loss of weight that commonly occurs in stored meat. The Commission is therefore right to assume that the meat removed from storage was not always weighed.

149 Furthermore, the Community rules provide that in the case of non-identified losses of beef, in excess of the tolerance limit of 0.6%, the Member States have to credit to the EAGGF a sum substantially higher than the sales value of the quantities lost. The Member States are therefore required to carry out rigorous controls to establish whether or not that limit has been exceeded.

150 The ex post documentary controls carried out by BALM in those cases in which it used representatives to carry out the control on removal from storage cannot compensate for the irregularities which may have occurred on removal from storage.

151 In that regard, the mere assertion that the purchasers made no complaint cannot be considered conclusive, if only because BALM's conditions of sale preclude the lodging of complaints after the event.

152 I therefore find that the German Government has provided only very partial proof that the deficiencies cited by the Commission in regard to the controls to be carried out when intervention beef is removed from storage do not exist.

VI - Final considerations

153 Contrary to what the German Government claims in its application, the summary report on which the Commission decision at issue was based does not describe the deficiencies identified as significant (beträchtlich).

154 Furthermore, the Commission stressed, during the procedure before the Court, that its decision was not based on the deficiencies identified at each of the three stages in the procedure viewed in isolation, but on an assessment of the procedure operated by the Federal Republic of Germany as a whole.

155 In my view, the arguments presented by the Federal Republic of Germany fail to demonstrate that this procedure does not contain defects which, viewed as a whole, may result in `the risk of minor losses to the EAGGF.' It is also clear from several documents in the file that the applicant undertook to remedy those shortcomings, proving that they did indeed exist.

156 In the light of the assessment of the controls carried out at each of the three stages, I therefore conclude that the applicant has not succeeded in demonstrating the inaccuracy of the Commission's findings.

157 I have, in addition, to note that, in its arguments, the German Government simply challenged the existence of the deficiencies on which the flat-rate reduction is based. It did not, however, put forward any plea designed to challenge the applicability of that reduction to the deficiencies identified.

158 In the context of an action for annulment, the Court of Justice merely considers the pleas of the applicant and is not, therefore, required to consider in this case whether the deficiencies identified justify application of the flat-rate reduction.

159 Consequently, I have to conclude that the application should be dismissed.

160 I wish, however, to add that, had the applicant raised the issue of whether the irregularities in question were of the kind that should trigger a flat-rate correction of 2%, in accordance with the scale of correction adopted by the Commission, my conclusion would have been the same. A correction of that nature applies `where the deficiency is limited to parts of the control system of lesser importance, or to the operation of controls which are not essential to the assurance of the regularity of the expenditure, such that it can reasonably be concluded that the risk of loss to the EAGGF was minor.' Now it can hardly be disputed that the deficiencies found to exist on conclusion of this analysis are likely to result in a minor risk of loss to the EAGGF.

161 Furthermore, the question whether the application, in such cases, of a correction amounting to 2% of the expenditure is incompatible with the principle of proportionality was not raised in the instant case.

VII - Conclusions

162 In the light of the foregoing I propose that the Court:

- dismiss the application; and

- order the Federal Republic of Germany to pay the costs.

(1) - OJ 1996 L 323, p. 26.

(2) - See Doc. VI/6335/95-EN of 27 March 1996 and Doc. VI/5112/96 of 14 June 1996.

(3) - OJ, English Special Edition 1970 (I), p. 218.

(4) - OJ 1990 L 94, p. 13.

(5) - OJ 1994 L 182, p. 45.

(6) - Case C-242/96 Italy v Commission [1998] ECR I-5683, paragraphs 58 and 59.

(7) - OJ 1989 L 91, p.5.

(8) - OJ 1993 L 225, p. 4.

(9) - See point 35 above.

(10) - See Case C-209/96 United Kingdom v Commission [1998] ECR I-5655, paragraph 43.

(11) - See, among others, Case 347/85 United Kingdom v Commission [1988] ECR 1749, paragraph 60.

(12) - Passage translated into French by the Advocate General.

(13) - See point 103 above.

(14) - Annex 7 to the application.

(15) - OJ 1991 L 17, p. 9.

(16) - See Council Regulation (EEC) No 3492/90 of 27 November 1990 laying down the factors to be taken into consideration in the annual accounts for the financing of intervention measures in the form of public storage by the European Agricultural Guidance and Guarantee Fund, Guarantee Section (OJ 1990 L 337, p. 3) and Commission Regulation (EEC) No 3597/90 of 12 December 1990 on the accounting rules for intervention measures involving the buying-in, storage and sale of agricultural products by intervention agencies (OJ 1990 L 350, p. 43).

(17) - OJ 1979 L 251, p. 12.