Special Report No 16/2000

on tendering procedures for service contracts under the Phare and Tacis programmes, together with the Commission's replies

(pursuant to Article 248(4), second subparagraph, of the EC Treaty)

(2000/C 350/01)

CONTENTS

>TABLE>

INTRODUCTION

1. The DAS audit work on transactions from the 1998 financial year revealed weaknesses in the implementation of some of the procedures for putting contracts out to tender under the Phare and Tacis programmes. At the same time, on the basis of a Financial Control Department report on the subject, the spokesperson for the discharge for financial year 1997 asked the Court of Auditors to examine the procedures for awarding contracts under these two programmes. Against this background an audit was performed on 120 technical assistance contracts (80 for the Phare programme and 40 for Tacis), most of which had been concluded between 1996 and 1998 following restricted invitations to tender(1).

2. The sample concerning the Phare programme covered 31 operations managed by staff at the Commission's headquarters and 49 operations effected under decentralised management procedures (DIS) in four Phare countries (Poland, Latvia, Bulgaria and Romania). The Tacis sample covered transactions effected by staff at the Commission's headquarters, mainly in Russia. Owing to the nature of the sample, the audit results cannot be used for statistical projections to the whole of the Phare and Tacis field. On the other hand, the sample does demonstrate the existence of real risks in the award of service contracts under the Phare and Tacis programmes. It also revealed the number of problems encountered (see the table), thus highlighting the weaknesses of the systems adopted by the Commission at the time.

3. In June 1999, the Commission introduced a new instruction manual for service, supply or works contracts concluded in connection with Community cooperation with non-member countries. The intention of the manual is to establish stricter and more consistent procedures for all external aid, without any distinction between programmes. If they are rigorously applied, the instructions can be expected to rectify various shortcomings that have been pointed out in this report and must be seen as a valuable aid in improving the management of aid to non-member countries.

THE LEGAL BASE AND THE REGULATIONS

4. The award of the contracts that were audited is for the most part governed by Articles 112 to 119 of the Financial Regulation and by the regulations that relate specifically to Phare (Article 7)(2) and Tacis (Article 6 of, and Annex III to the Regulation)(3). The Financial Regulation (Article 117) stipulates that the tender selected must be economically the most advantageous, whilst authorising (Article 118(1)) the use of restricted invitations to tender for awarding contracts for services and for technical cooperation.

5. The Phare Regulation specifies (Article 7) that, in cases involving more than EUR 50000, participation in invitations to tender and contracts must be open on equal terms to all natural and legal persons of the Member States and the countries of the Phare group. The practical rules for awarding these contracts are laid down in the manual of decentralised management procedures (DIS).

6. For its part, the Tacis Regulation lays down that:

(a) the widest possible participation in invitations to tender for supplies, works and services under equal conditions must be ensured;

(b) the Commission must apply the selection criteria with due regard to transparency and rigour and at the same time try to diversify the selection;

(c) on the widest possible basis and using the most modern means of communication, any party that may be interested in tendering should be encouraged to express their interest.

7. Article 6 of the same Regulation stipulates that, as of the amount of EUR 200000, service contracts must, as a rule, be awarded by restricted invitations to tender.

THE MAIN STAGES OF THE RESTRICTED INVITATION TO TENDER

8. For both Phare and Tacis, there are four main stages in the restricted procedures for putting contracts up for tender:

(a) the call for expressions of interest in the proposed project;

(b) the drawing-up of a shortlist of undertakings that are to be invited to participate;

(c) the invitation to tender and notification of the detailed rules for submitting tenders;

(d) the evaluation of the tenders submitted and subsequent selection of the best tender.

The call for expressions of interest and the drawing-up of shortlists are vital, as they ensure that the principles of equality of treatment for the purposes of participation in public contracts are respected (Article 114 of the Financial Regulation). Similarly, the selection of the most economically advantageous tender relies on the criteria for evaluating the tenders (Article 117 of the Financial Regulation) as notified at the time of the invitation to participate.

9. Where operations are managed centrally, as is the case for the Tacis programmes and the Phare multicountry measures and the measures that were being carried out at the time in the Baltic States, employees at the Commission's headquarters are directly responsible for the procedure. Where operations are conducted under the decentralised management system, the DIS procedures are essentially carried out by the authorities in the beneficiary country, usually with the help of Western technical assistants and under the supervision of the Commission delegation on the spot. In all cases, the Commission retains overall responsibility for the correct use of the budget appropriations.

Calls for expressions of interest

10. This stage, which aims to ensure that the range of competition is the widest possible, consists in informing any interested undertaking or organisation of the forthcoming award of a contract. This information is circulated via the Official Journal of the European Communities or the Internet (since October 1996 for service contracts exceeding EUR 200000).

The drawing-up of shortlists

11. On the basis of the expressions of interest and their registers of contractors, the Commission staff or the PMUs(4) draw up a shortlist of undertakings which includes only those considered to be most able to carry out the planned operation. The factors borne in mind when drawing up the shortlist are recorded in a report which is included in the file for the invitation to tender.

12. In order to ensure that there is adequate competition, the two programmes have laid down a minimum for the number of undertakings that are to appear on the shortlist: seven for contracts not exceeding EUR 1 million and 10 for other contracts. Paragraph 11 of Annex II to the Tacis Regulation also states that in the case of highly complex projects it may be suggested that undertakings be grouped in consortia.

Invitations to participate

13. The undertakings chosen at the end of the preselection process are invited to submit a tender for services. For this purpose, they receive a complete file containing the following documents:

(a) the list of undertakings invited to participate;

(b) the instructions to tenderers;

(c) the rules applicable to tenders and to the award of service contracts financed by Phare and Tacis;

(d) the standard contract and its annexes, including the specifications for the particular contract concerned.

14. The instructions specify the structure, content, conditions and deadlines for the submission of tenders and the criteria for the award of the contract. These are the factors on which the tenderer bases his bid.

15. The selection process always has two stages: a technical evaluation, followed by a financial evaluation, which account for 70 % and 30 % respectively of the final appraisal.

The evaluation of tenders

The evaluation committee

16. The bids submitted by the tenderers are evaluated by a special committee. Since October 1996 the evaluation committees for contracts awarded on the direct authority of the Commission have comprised a (non-voting) chairman, who is an established member of the Commission staff, and five evaluators (two Commission representatives, one representative of the final beneficiary, two outside experts or two experts belonging to the Commission staff). For the decentralised Phare contracts, the local authorising officer determines the composition of the committee, which may include an outside expert.

17. When the tenders are opened, the evaluation committee checks that each tender was indeed received within the prescribed time limits and that the technical and financial proposals are separate. It then proceeds with the technical and financial evaluations, checking that the documentation was duly submitted in accordance with the prescribed rules. At each of these stages the Committee draws up a report on its work.

Technical evaluation

18. The first part of the actual evaluation is the appraisal of the technical proposals and this usually consists of two stages:

(a) scrutiny of the written proposals;

(b) an interview with the main parties involved (head of project, long-term experts, etc.).

19. The technical tenders are evaluated in the light of various parameters, which, as a rule, are set out in the instructions to tenderers. The evaluators take into account the proposed organisation, the tenderer's experience and the quality of the proposed staff, plus the use of local experts. The purpose of this latter point is, in principle, to encourage the transfer of know-how and to ensure that the results achieved are sustainable in the long term. Only undertakings which score over 65 % in the technical evaluation then go on to the financial evaluation (see paragraph 15).

Financial evaluation

20. The instructions to tenderers always state that the proposed fees must be within reasonable market rates and that the total amount of the tender may not exceed a given budget for each invitation to tender.

21. For contracts under the Tacis programme, the information on the financial evaluation does not contain any further details, whereas for Phare contracts the Commission's internal instructions (4 October 1996) require the method used for the financial evaluation to be specified. For the latter, one of the following two methods is usually applied:

(a) comparison of the total prices, excluding reimbursable expenses;

(b) comparison of the unit prices (total price less reimbursable expenses divided by the proposed number of man-days).

22. The mark awarded for the financial proposal is added to the mark for the technical part in order to obtain the final mark. The tender that achieves the highest mark overall is then regarded as economically the most advantageous and, in principle, gains the contract.

OBSERVATIONS ON RESTRICTED INVITATIONS TO TENDER

23. The audits revealed difficulties at various stages of the invitation-to-tender procedure. The most serious problems, however, concerned the evaluation of the tenders and the management of the files. These shortcomings were found in 47 of the 80 Phare files and 21 of the 40 Tacis files examined.

24. Of the 68 cases in which the problems encountered were considered to be serious (see the table) there were eight where the problem had a direct effect on the final award of the contract and 12 where it might have affected the award of the contract. Due to the absence of supporting documents, it is not possible to give an opinion in the latter cases.

Table

Problems considered to be serious in the sample of 120 contracts examined

>TABLE>

Shortlists

Quality of the preselection

25. With the exception of four Phare contracts for which the documentation could not be produced, all the contracts examined were duly advertised as required.

26. There was considerable variety in the form and the content of the preselection reports, depending on the programme and the manager. Under the decentralised management for Phare, an evaluation chart is used consisting of predetermined criteria which result in an overall mark. Only those undertakings that have been awarded a mark higher than a set minimum of 65 % are allowed to go on to the next stage. The most recent version of the decentralised management procedures came into force in September 1997, since when the preselection has been the responsibility of a special committee made up of three persons designated by the local authorising officer.

27. In the case of the Tacis programme, the shortlist is essentially drawn up on the basis of an appraisal by the Commission manager, endorsed by the authorising officer.

28. Thirty-four Phare and seven Tacis files did not contain documentation to show that the shortlist had been drawn up in a way which ensured that there was effective competition.

29. Irrespective of the number of undertakings invited to tender, the proportion that goes on to the technical evaluation stage is an indicator of the interest in the contract and, therefore, of the level of competition achieved through the shortlist. For 13 Phare contracts and three Tacis contracts (Annex I), only one tenderer passed the technical selection and was simultaneously awarded the contract.

Reduced competition

30. In fact, the setting of minimum levels for the number of undertakings to be included in shortlists had little effect, because the formation of consortia and subcontracting between undertakings on the same shortlist were authorised very often, thereby weakening the competition between tenderers (see box No 1). The instruction manual mentioned in paragraph 3 now prohibits groups of this kind.

31. Thus, in 34 of the 80 Phare invitations to tender and 31 of the 40 Tacis invitations to tender examined, undertakings on the shortlist were grouped together. Furthermore, in 38 Phare cases and 11 Tacis cases, the number of tenders received was less than half the number of invitations to tender that had been sent, partly because of grouping (Annex II). Moreover in one Tacis case there were questionable aspects to the manner in which one tenderer went about forming a consortium (see box No 1, paragraph 35).

Box No 1

Phare

32.

105:EUR 1 million. There were 11 undertakings on the shortlist for the award of a contract in the energy sector in Romania. As a result of grouping, only three tenders were submitted.

33.

110:EUR 2,3 million. The instructions to tenderers for an institutional development operation in Romania, on the one hand, stipulated that the undertakings on the shortlist must form consortia and, on the other, expressed the wish that a context of normal competition could be maintained. The outcome was that there were only two tenders combining six associates out of the 16 that had been invited to tender.

Tacis

34.

24: EUR 2 million. Although eight undertakings were initially invited to tender, only two bids were received for a contract in the transport sector. These two tenders involved five of the undertakings from the preselection.

35.

27:EUR 2,5 million. For the award of a contract relating to the design of a system of payment by bank card in Russia, one of the tenderers was approached twice (first directly and then via an intermediary) to participate in a consortium with one of the other preselected undertakings. The intermediary was confident of being able to win the contract if his proposal was accepted: participation in the consortium and payment of a commission. The proposal was rejected and the tenderer who had thus been contacted informed Commission staff of what had happened. Despite the details of this case having been corroborated by two Commission officials, the contract was awarded to the consortium which included the offending undertaking.

36. The Tacis evaluation reports clearly state which of the partners in the consortium were originally on the shortlist. The Phare files, however, do not always provide this information.

The information supplied to tenderers

Rules for the formation of consortia

37. As a general rule the instructions stipulated that consortia and subcontracting contracts were permitted between undertakings on the shortlist, but it also happened that such associations were not allowed (four Phare cases, six Tacis cases). Collaboration is possible under various rules, but as no justification is ever given for the option chosen, it is difficult to assess the reasons for it.

Mention of the financial evaluation criteria

38. No precise method is specified for Tacis contracts, a fact which does not comply with the provisions of Article 117 of the Financial Regulation. Furthermore, it is not in accordance with best practice, which requires that the award criteria be stated in the contract specifications or the tender notice whenever the contract is to be awarded to the economically most advantageous tender (as opposed to an award based on other criteria).

39. For 34 of the 80 Phare files, this information was not given. Furthermore, for 12 Phare files the Commission's files did not contain the instructions to tenderers.

Use of outside experts

40. Outside experts are involved, on the one hand, in drawing up the contract specifications and, on the other, in evaluating the tenders. They are recruited on framework contracts which cover most of the sectors of assistance (13 for Phare and six for Tacis) (see paragraph 70). There are two possibilities:

(a) the Commission manager proposes an expert whom he knows. The administrative situation is subsequently regularised by integrating this expert, after negotiation, into the framework-contract teams;

(b) the Commission manager submits the desired profile to the contractor which proposes one of the experts on its lists.

The procedure still lacks transparency and is protected against possible confusions of interest by a formal system of sworn declarations (see box No 2).

41. As of 2000, the implementation of new framework contracts (at least three per sector of assistance) should make it possible to improve the situation. The managers should no longer have to propose candidates and the desired profiles are expected to be submitted in parallel to three framework contract holders in the sector concerned.

42. In order to increase the effectiveness of the systems, the Commission should set up, for the external aid programmes, a database containing the main data on the experts. A tool of this kind would make it possible:

(a) to reduce the risks of confusions of interest from the very outset. In two cases, beneficiaries of framework contracts officially expressed their interest in service contracts in their sector of activity, even though their contracts prohibit them from doing so (

35: EUR 0,9 million;

51:EUR 2,8 million);

(b) to improve circulation of experts' profiles, and their track records in particular, at the Commission. Such circulation is essential, in view of the high turnover of personnel at the Commission and the widespread use of experts. Since October 1996, more than 2000 experts have been consulted through Phare framework contracts.

43. Confusions of interest were detected in seven Phare and two Tacis cases. The bids presented under these circumstances were in fact rejected on three occasions in the case of Phare and once for Tacis, but not in the other cases (see box No 2).

Box No 2

Phare

44.

101:EUR 1,5 million. The person who drew up the contract specifications for a pilot project for the regional land register in Bulgaria was proposed as the long-term expert in the bid submitted by one of the tenderers. This undertaking, which thus had a crucial advantage, succeeded in winning the contract. This confusion of interests was not mentioned in the evaluation committee's minutes, but was mentioned in a follow-up report that was submitted subsequently. Not only was the contract signed with the offending undertaking but the expert also remained on the technical assistance team.

Tacis

45.

80:EUR 2,9 million. In the case of one Tacis contract relating to the environment sector, a person consulted by the Commission when the specifications for the project were being drawn up also appeared as the coordinator in one of the tenders submitted for the same project. The confusion of interests did not result in the exclusion of either the expert or the undertaking concerned; indeed, the undertaking won the contract, having been awarded a very high mark for the technical part.

Evaluation of the tenders

Operation of the evaluation committees

46. Even though the members of the Committee are required to attend the evaluation meetings, five Phare files and one Tacis file showed that the evaluators had not attended all the meetings. In one case, the evaluator who was absent was simply treated as an observer and his evaluation was not taken into account. Some of the solutions accepted by the Committee in the case of Phare were debatable: in particular, the replacement of evaluators during the procedure, the awarding of marks for stages of the evaluation in which the evaluators in question had not participated and the signing by those absent of attendance sheets for meetings which they did not attend (see box No 3).

Box No 3

Phare

47.

104: EUR 0,9 million. In one case relating to the tourism sector in Poland, the marks given by the evaluator who attended some of the evaluation phases were simply repeated for the phases where he was absent. Furthermore, these absences did not prevent the evaluator from signing the attendance lists for meetings which he did not attend.

48.

29:EUR 1,8 million. For one contract in the marketing of agricultural products sector, the independent expert submitted a report to the Commission, calling in question the reliability, competence and independence of the other members of the Committee during the evaluation of the tenders. His conclusion was that the evaluation had not been taken seriously.

49.

109:EUR 1,4 million. For one training project, one of the external evaluators stated in his report that other Committee members were frequently absent, to the extent that at one interview he was on his own. Nevertheless, a mark was given by every evaluator for every interview.

Marks awarded to the technical proposals

50. In view of its weight (70 %), the technical evaluation is decisive and any significant difference between the tenderers as regards the points awarded is a serious handicap in the later stages. It must, therefore, be considered very carefully: in all the files examined, there were only three contracts that were awarded to an undertaking which did not occupy the top position at the end of the technical evaluation.

51. In the case of 16 Phare and 12 Tacis files, the marks awarded by the local assessors differed greatly from those given by the other members of the Committee and were the cause of significant differences between the undertakings evaluated. In two Phare cases and one Tacis one, the divergences led the Commission to cancel the tender evaluation. In other cases, the results were accepted and the procedure continued. The reasons for these differences in attitude should be explained carefully when the decision is tipped in the direction of a particular tenderer precisely because of differences of this type (three Phare and three Tacis cases).

Box No 4

Phare

52.

PL05: EUR 0,4 million. The marks awarded by one member of the evaluation committee swung the award of a contract in Poland (technical assistance for privatisation) in favour of one of the tenderers.

Tacis

53.

94: EUR 3,9 million. As a result of overmarking by the only Russian expert, one Tacis contract concerning the development of human resources was awarded to the undertaking that came second at the end of the technical evaluation. The difference between the marks awarded to the top two undertakings by the expert was four times greater than the difference between the four other assessors' evaluations of the same tenderers. If this expert's marks had not been taken into account, the contract would have been awarded to another undertaking.

54. There is no rule to distinguish cases where the procedure must be terminated from those where it may be continued. Under such circumstances it would be preferable if the marks at the extreme ends of the scale were excluded when the average is being calculated (e.g. if they deviate from the average by more than a predetermined amount), or, if appropriate, that the evaluation be taken over by another committee.

Interviews

55. The interviews are part of the technical evaluation and the points awarded count towards the final technical mark. In the case of Phare, the written tender usually accounts for 80 % of the technical evaluation and the interviews for 20 %. On occasion, however, the interviews either accounted for more than 20 % (12 Phare files), or the evaluation committee did not arrange an interview (18 Phare cases). In 16 Phare cases and all the Tacis ones, the proportion applicable to the interview was not specified and adjustment was a matter for overall discretionary appraisal.

56. There is nothing in the Phare files to explain the choice of options. Moreover, by omitting to specify in the instructions to tenderers whether interviews would take place and by not drawing attention to their importance, the Commission's departments did not follow the rules provided in Article 117 of the Financial Regulation (see paragraph 4). Finally, contrary to what is laid down in the instructions to tenderers, a selection has very often already been made on the basis of the written proposals. Only undertakings which obtain more than 65 % from the technical review of their files are invited for interview. However, in reality, many tenderers have had their technical mark raised after the interviews. Given that this preliminary filter is never mentioned, the information sent to tenderers is incomplete.

Financial evaluation methods

57. It was found from examining the files that 11 different methods had been used in evaluating the financial proposals for Phare and six for Tacis (Annex IV). These methods can be divided into two groups: in the first, the comparisons are based on the overall amounts, whereas in the second it is the average unit costs that are compared. Some of the methods applied in the second group produce effects which go against the policies set out elsewhere by the Commission.

58. This is the case where, in order to calculate the average cost of the experts, the total cost of experts, both local and European, is divided only by the services rendered by the European Union experts (man-days). This method gives the advantage to tenders which favour hiring European experts, to the detriment of those which employ local experts. Such a step is in conflict with the intentions expressed in the instructions issued to tenderers, and in the Commission's internal instructions (4 October 1996), which seek to promote the use of local experts, with the aim of ensuring that the results achieved have long-term sustainability (see paragraph 19).

59. In the sample examined, six Phare and nine Tacis contracts were affected by the non-inclusion of experts from CEEC countries. In only one case (

62: EUR 4,8 Million) did a Commission official ask for the basic calculation to be amended so as to place all tenderers on an equal footing.

Changes in the methods of financial evaluation

60. While the evaluation criteria were indeed sent to tenderers in the case of the Phare programme, the evaluation committee nevertheless applied different criteria sometimes. Thus, the evaluations of 12 of the 80 Phare contracts examined were carried out under questionable conditions. This practice may mislead tenderers, who construct their tenders according to the criteria issued to them. In two of the 12 contracts involved, the contract would have been awarded to another tenderer if the criteria had been adhered to. The absence of financial tenders in two other files made it impossible to say whether the contracts would have been awarded to the same tenderers if the method proposed initially had been applied (see box No 5).

Box No 5

Phare

61.

33: EUR 11,7 million. The instructions issued to tenderers specified that the financial proposals would be evaluated on the basis of the average unit cost of the experts. This meant the total cost, less reimbursable expenses, divided by the experts' services calculated in terms of man-days. These instructions also specified (Section C.2.1) that an incorrect allocation between direct costs and reimbursable expenses could lead to the rejection of the tender and that any expenditure relating to support for the project must be entered under direct costs. None of these provisions was applied.

The financial evaluations were thus made on the basis of overall costs excluding reimbursable expenses, without considering the number of experts, and therefore, prices being equal, conferring an advantage on the tenders that used fewer experts. Furthermore, the tender selected included, among the reimbursable expenses, various support measures, in particular EUR 288000 for local offices, which should have been included under the direct costs. The proposal in question should have been rejected as soon as the tenders were opened, on the grounds of incorrect cost allocation. Lastly, examination of the financial proposals shows that the composition of the reimbursable expenses varied from tender to tender and no corrections were made before the financial evaluation in order to ensure that the data being compared was of the same type. It thus emerges that, if the evaluation criteria issued to the tenderers had been applied, the undertaking chosen would not have been selected since it would have been placed lowest in the rankings.

Calculation data

- Evaluation which did not respect the advertised criteria:

>TABLE>

- Evaluation respecting the advertised criteria:

>TABLE>

62.

56: EUR 4 million. The instructions issued to the tenderers for a technical assistance contract in Latvia informed them that the financial proposals would be evaluated on the basis of the average unit cost of the experts. It was therefore in the tenderers' interest to keep this average cost as low as possible, in particular by making greater use of local experts whose fees were, as a rule, lower. In reality, the result of the evaluation was distorted by the way in which the calculation data were handled. The total cost for experts, both European and local, was only divided by the time required by the European experts to perform their services. The tender chosen was the one which favoured the use of (costly) European experts, to the detriment of those which used a larger number of (cheaper) local experts. Had the data for the calculation not been revised, the contract would have been awarded to a different undertaking.

Calculation data

- Evaluation which did not respect the advertised criteria:

>TABLE>

- Evaluation respected the advertised criteria:

>TABLE>

63. Similarly, in three Tacis cases, the information was not sent to the tenderers, but contradictions were found between the criteria initially agreed by the evaluation committee and the way in which they were actually applied. However, the changes did not affect the results of the evaluation.

File management

64. Even though the Tacis files are more complete and their quality is better than that of the Phare files, the management of files by the Commission staff reveals three types of shortcoming:

(a) documents dispersed;

(b) documentation incomplete or mislaid;

(c) difficulty, or even impossibility, of identifying the manager responsible for the project (because certain non-permanent managers left the Commission).

65. In the case of 27 Phare and 15 Tacis files, it is not possible to assess whether the procedures for awarding contracts were entirely regular, because the tenders that were eliminated, or the evaluation reports, are missing. In addition, the information necessary for a proper understanding of the course that the procedures took is not always available. Some important items of correspondence were not placed in the files and reports drawn up by the expert evaluators were not always placed on file systematically, even in the case of headquarters management.

66. The management of the Tacis programme files was entrusted to a technical assistance office (TAO), whereas that of the Phare programme was carried out by the local contracting authorities and the Commission. After having been managed centrally, the Phare files were transferred to the delegations as from the second half of 1998 for the national programmes and to the SCR for the multinational programmes.

67. The conditions under which these transfers were carried out were not satisfactory. Firstly, the majority of the files sent to the delegations proved to be incomplete. These files were then transferred, but in no real order and with no accurate notification of their contents. At the time of the audit visits, in June 1999, the delegations had not yet received guidelines on filing (how to keep and structure files).

68. As regards Tacis, it is rather unwise to entrust a private operator with information as sensitive as that connected with the awarding of public contracts for which undertakings in the same group as the operator in question may tender.

69. This situation is the result of administrative weaknesses which have been highlighted many times in the Court's reports: very high rate of turnover among managers, combined with a lack of clear rules on the keeping of files and their structure and storage. These management defects mean that it is not possible to guarantee continuity of service in the managing departments (institutional memory). In the case of the contract documentation, for which the Commission may be held responsible, greater administrative discipline is called for.

OBSERVATIONS CONCERNING FRAMEWORK CONTRACTS

70. In order to speed up the implementation of the Phare and Tacis programmes and reduce the number of invitation-to-tender procedures (Annex V), the Commission concluded framework contracts in various sectors (13 for Phare and six for Tacis). The framework contracts were awarded following open invitations to tender. The files for the Phare contracts examined were mislaid by Commission staff.

71. The Commission insists that framework contracts be used for any service contract involving an amount of less than EUR 300000 and for a duration of less than 18 months. This form of contract is used, in particular, for recruiting experts to draw up the contract specifications and evaluate the tenders (see paragraph 40).

72. In the case of Phare, the contracts were signed in September and October 1996. At the time, their duration was supposed to be limited to two years, with no extension possible. Subsequently, all these contracts were extended three times, taking their overall duration to almost three and a half years. The total value of the contracts in question was EUR 106 million (Annex V).

73. In the case of Tacis, six framework contracts were signed in May 1998. Of the 45 tenders received for the six sectors concerned, 26 were rejected because they did not involve the minimum number of experts required (70).

74. The condition concerning the minimum number of experts, which seriously reduced competition, appears debatable. In fact, the Commission managers themselves often propose experts, who are then entered on the contractor's list ex post facto(see paragraph 40). In the particular case of the human resources sector nine of the 10 bids received were excluded for that very reason, thereby eliminating any real competition. For four out of the six sectors only one undertaking reached the financial evaluation stage, and won the contract at the same time.

75. Lastly, the new principles governing the use of technical assistance offices (TAOs)(5), which were adopted in June 1999, specify that contracts concerning TAOs must first be submitted for the opinion of the Advisory Committee on Procurements and Contracts (ACPC). As with the TAOs, the new framework contracts too should, in future, be submitted to the ACPC.

CONCLUSIONS AND RECOMMENDATIONS

76. The audit revealed various problems in the awarding of service contracts. The most important of them concern the evaluation of the tenders and the existence of supporting documents. These findings concern more than half the Phare and Tacis files that were examined (see the table).

Competition between tenderers

77. Although the use of the restricted invitation-to-tender procedure does not necessarily entail any restriction of competition, some decisions nevertheless effectively limit competition, for example, the formation of consortia and the use of authorised subcontracting among enterprises included on the same short- list (see paragraphs 25 to 36).

78. Partnerships between undertakings on the same shortlist are no longer permitted, in order to avoid agreements whose main effect is to weaken competition. However, it might be possible for partnerships with other undertakings to be approved, especially where they have expressed an interest in the project but have not been included on the shortlist.

Transparency of procedures

79. The instructions sent to tenderers do not always specify or explain the criteria and procedures governing the award of the contract (Phare), or they omit them (Tacis). The omissions concern, firstly, interviews, which may be arranged (or not), and whose importance varies from one contract to another (see paragraphs 55 and 56). They give particular cause for concern when they concern the criteria used for the financial evaluation (see paragraphs 38 and 39). Most of the files do not include the financial bids that were eliminated, which makes verification of the procedure difficult (see paragraphs 64 to 69).

80. In such a context, the procedures for awarding contracts are open to question when the evaluations take place under procedures that differ from those notified to the tenderers (see paragraphs 57 to 63).

81. If the Commission wishes to meet the standards of transparency it has set itself for the awarding of contracts, it should in future:

(a) draw the tenderers' attention to the importance attached to the interviews and notify them of the methods adopted for the technical and financial evaluations;

(b) ensure that the evaluation committees do not deviate from the evaluation criteria and rules issued to the tenderers, and that they always take care to justify all their decisions and initiatives;

(c) establish a standard filing system in which all the information concerning an invitation to tender (tenders received, detailed reports by the evaluating experts, correspondence, documents relating to the negotiations) is collected in one case file;

(d) send clear, standard filing instructions to all its delegations and check that they are actually applied;

(e) in the interests of transparency, record, accurately and comprehensively, the follow-up given to complaints and irregularities.

Effectiveness of the procedures

82. The continuity of public service and adequate knowledge of procedures are not effectively provided, neither by means of personnel (substantial staff turnover) or files (often incomplete), nor by means of regulations (lacking clarity) (see paragraphs 4 to 7, 64 to 69).

Responsibilities of the parties involved in the procedures

83. Absences on the part of evaluators and marks that vary excessively may indicate a lack of rigour and independence on the part of the evaluators concerned (see paragraphs 40 to 45, 50 to 54).

84. It is suggested that the organisational framework be improved for all those involved and that any risk of a confusion of roles be eliminated at the outset. This requires, in particular, the creation of a database containing details of all the outside experts and their performance in relation to their successive contracts. The effects of any anomalies that are noted in the marks awarded could be mitigated by eliminating some of the extreme notes when the average is calculated. Moreover, the working of the evaluation committees might be improved if an agent of the Commission were present in the evaluation committees convened on the initiative of the decentralised contracting authorities (see paragraph 16).

General conclusion

85. While open invitations to tender are the surest way of guaranteeing a broad range of tenders for contracts, making their application more widespread would increase both the cost of operations (personnel, experts, costs) and the time required. Furthermore, it is the discipline with which the preselection, information and, above all, evaluation procedures are carried out that should receive particular attention, rather than the form of the invitation to tender.

86. There can be no justification for the existence of so many differences between Phare and Tacis as regards the procedures followed in awarding service contracts. In this respect, harmonising the procedures for awarding service contracts within the Commission would be a highly effective means of introducing the discipline which has, until now, been lacking. On the same point, creating the SCR and standardising the procedures for contracts concluded within the framework of cooperation for the benefit of third countries is only a first step: the desired harmonisation should be carried out at the level of the Financial Regulation applicable to the Community budget. The problems that were found only serve to emphasise the urgency of this harmonisation and the need to draw up simpler and lasting rules. Furthermore turnover among the staff allocated to tasks which are essentially civil service duties should be reduced.

The report was adopted by the Court of Auditors in Luxembourg at the meeting of 12 and 13 July 2000.

For the Court of Auditors

Jan O. Karlsson

President

(1) Five of the Tacis contracts examined for the DAS were earlier than 1996.

(2) Council Regulation (EEC) No 3906/89 of 18 December 1989 (OJ L 375, 23.12.1989), as last amended by Regulation (EC) No 1268/1999 of 21 June 1999 (OJ L 161, 26.6.1999).

(3) Council Regulation (EC) No 1279/96 of 25 June 1996, OJ L 165, 4.7.1996.

(4) Programme Management Units.

(5) Communication from Mr Liikanen to the Commission, SEC(1999) 981, 23.6.1999.

Annex I

Phare contracts for which only one undertaking passed the technical selection stage

>TABLE>

Tacis contracts for which only one undertaking passed the technical selection stage

>TABLE>

Annex II

Changes in the average number of tenders at different stages of the procedure

>TABLE>

Annex III

Infringements of the rule limiting the size of Tacis contracts (1996, 1997 and 1998)

>TABLE>

Annex IV

Phare financial evaluation methods

>TABLE>

Tacis financial evaluation methods

>TABLE>

Annex V

List of framework contracts

>TABLE>

The Commission's replies

INTRODUCTION

1. The Court's report looks at an important area of the Community's external aid management. The difficulties arising from the expansion of the two programmes, Phare and Tacis, over the period 1990 to 1997 and the plethora of procedures applicable to Commission programmes in different geographical regions led at the end of 1998 to substantial changes in both the structure of the Commission departments concerned and basic procedures, especially with a view to the reforms announced in May 2000.

The objectives of the reform are results-oriented. The aim is to improve the quality of operations through better programming and to reduce the time taken to implement projects. Many evaluations have confirmed that the main obstacles to a speedy and successful implementation of these programmes are a lack of human resources and overly complex procedures. The reform will try to find realistic solutions to these problems.

The Commission attaches great importance to compliance with contract-award procedures, and many measures have been taken under the Phare and Tacis programmes, especially since 1997, to bring about more rigorous and more transparent competition.

The Phare programme, most of whose recipients are candidate countries for accession, is essentially decentralised and devolved, while the Tacis programme, the recipients of which are States of the former USSR, is completely centralised as the recipients are not preparing for accession. Under Phare, the national authorities generally assume responsibility for the procedures, issue and conduct invitations to tender and sign contracts, while in the case of Tacis contracts, it is the Commission departments which carry out all the procedures.

The Phare programme is much influenced by the national characteristics of the applicant countries; the level of maturity and the administrative culture varies greatly from one country to another. Under these conditions, the controls on procedures cannot be completely standardised and faultless in every case, whatever the efforts made by the Commission to train the various contracting authorities concerned and to publish guides and manuals (in particular the decentralised implementation system manual - DIS), designed to ensure consistent and clear procedures.

For Tacis the Commission has made a considerable effort to develop clear and exhaustive management tools, (for example the Tacis manual of operations). Some of the procedures used for a time have been abandoned and the Commission has approved standard service contracts and contract-award procedures that incorporate a great many changes and improvements.

OBSERVATIONS ON RESTRICTED INVITATIONS TO TENDER

23. The Court claims to have found shortcomings in 47 of the 80 Phare files and 21 of the 40 Tacis files that were examined. The Commission notes that in 21 of the Phare cases and 15 of the Tacis cases, the shortcomings concern only the file-keeping.

24. Without casting doubt on the substance of the Court's arguments, the Commission does not share its conclusions regarding three of the eight cases mentioned.

Shortlists

Quality of the preselection

28. With regard to Phare, the shortlist at that time was taken directly from an internal database, the central consultant register (CCR). All the firms interested in taking part in invitations to tender could get themselves entered in this database by providing proof of their experience in a given field. When issuing an invitation to tender the contracting authorities used this list to select the tenderers. This is why no detailed supporting arguments concerning the composition of the shortlists were included in the files.

Under the Tacis programme the drawing-up of shortlists was subject to controls at several levels, in particular by the head of the geographical unit concerned, the head of the horizontal unit responsible for the control, the director and, if necessary, the deputy Director-General, depending on the value of the tender. Each shortlist was normally accompanied by detailed information, including the complete list of the expressions of interest received from operators for the project concerned, the grounds for the proposed selection, the list and the value of the contracts obtained previously by the operators selected, and the projects for which these operators had been registered on the shortlist. The process of controlling the shortlists was accompanied by guarantees as to transparency. While it is true that certain documents sometimes do not appear in the 1994 to 1998 files, owing to successive removals as well as structural changes, files have been systematically kept from the end of 1998.

29. The first step in the selection procedure is to check that the firms have the economic and financial capacity required, plus broad experience and suitable references. The initial step does not involve any a priori regarding the number and quality of tenders to be submitted. The quality depends on the effort the firm puts into its tender, the attention it pays to what the project is meant to do on the basis of the tender specifications and other technical factors.

The fact that only a small proportion of firms pass the technical evaluation stage shows that there is a detailed and rigorous examination of the qualitative aspects of the tender by the evaluation committee. Indeed, the ultimate purpose of an invitation to tender is to select one bid among those submitted.

Reduced competition

30. The setting up of consortia between tenderers is a practice which was encouraged explicitly under the Phare and Tacis programmes in order to encourage the submission of tenders with a broad European base, to improve technical capacity and the transfer of tenderers' know-how, and to cope with the specific features of the environment in the recipient countries and the scale of local problems. This complementarity was seen as an important advantage and arose from a considered policy agreed with the Council. This provision was moreover legal and part of the Tacis Regulation. As implementation of the programmes advanced, and in the light of the evolving situation in our partner countries, this policy was scaled down, then abandoned.

31. The tables annexed to the Court's report show that competition in invitations to tender was broad-based and real (on average for Phare, 27,2 expressions of interest, 9,5 firms on the shortlist, 4,7 bids received and 2,4 bids making it through to the financial evaluation stage and, for Tacis, 27,6 expressions of interest, 8,4 firms on the shortlist, 5,2 bids received and 2,7 bids making it through to the financial evaluation stage).

Box 1

Phare

32. It is quite usual in an invitation to tender for services, whatever the programme or region in question (in this case, the Romanian contracting authority), for the number of tenders received to be more or less the same (around 25 %). Moreover, three of the tenders involved nine firms.

33. In this case, the Romanian contracting authority considered that the very specific field covered by this project and the complexity of the terms of reference required the setting up of a consortium between firms of at least two Member States. The initial number of firms invited (16) gave a wide base for competition and six firms finally took part in the two tenders submitted.

Tacis

34. As mentioned earlier, this practice was in complete conformity with the provisions of the Regulation and the two bids received involved five of the eight firms initially invited.

35. In this case, the charge arises from the suspicion that the representative of the recipient country was likely to be biased in the evaluation. The Commission was informed of these allegations and the evaluation was the subject of close scrutiny.

Owing to the number of members (five) of the Tacis evaluation committees, it is very difficult for only one member of the committee to skew an evaluation. At the end of the evaluation, it was clear that the vote of the recipient's representative had no influence on the final choice in the way suspected by the Court, in so far as that he granted his highest mark to the complainant, and the result would have been the same even if this mark had not been taken into account. Since the evaluation did not show evidence of any irregularity, the Commission awarded the contract.

The information supplied to the tenderers

Rules for the formation of consortia

37. Under both Phare and Tacis tenderers are always notified if the creation of consortia is not authorised. The choice depends on the technical nature of the project, which is described in detail in the terms of reference.

Mention of the financial evaluation criteria

38. The list of the general evaluation criteria was included in the instructions to tenderers that were always attached to the Tacis tender dossier.

39. The instructions to tenderers are normally included in the tender dossier sent to tenderers.

Use of outside experts

40. It is true that the routine use of external experts to carry out evaluations of invitations to tender for public contracts can present risks. The Commission considers it regrettable that, in the absence of sufficient statutory personnel and of high-level technical expertise that is only available on the market, it has no other choice but to resort to the private sector to carry out such tasks. However, it should be stressed that the external experts do not have decision-making powers and that they only submit recommendations to the contracting authority via the conclusions of the jury.

While these potential risks exist, the experts' selection is completely transparent. Framework contracts were concluded following an international invitation to tender with firms that were selected by a very competitive process. These firms are officially banned from taking part in invitations to tender in fields falling within their framework contract. The sworn statements made by the evaluators render them personally accountable, in accordance with modern principles of sound management.

42. The Commission regrets not being able to implement the Court's suggestion concerning the dissemination of data on experts giving their profiles and track records, owing to a lack of human and material resources (in particular computer resources).

43. In the case of Phare, the Commission acknowledges that errors were made by the local contracting authorities, affecting four cases in two countries.

However, the Commission does not believe there was any confusion of interest in the case of Tacis (see paragraph 45).

Box 2

Phare

44. In this case, it is true that the Bulgarian contracting authority should have prohibited the firm from proposing the expert in question, or even eliminated this firm's tender.

Tacis

45. In this case, the Commission employed three people in no way involved in the tender to draw up the specifications. The fourth person, to whom the Court refers, did not take part in the drafting of the specifications and played no role in the invitation-to-tender procedure. The contract was therefore awarded in a completely normal manner.

Evaluation of the tenders

Operation of the evaluation committees

46. In the case of Phare, it happened that evaluators were prevented at the last minute from taking part in meetings for reasons of force majeure (illness, travel or other impediments). Since the objective of the contracting authority was to ensure a fair evaluation by competent people, it accepted their replacement by other competent evaluators. The case where it is alleged that an evaluator signed the list of participants for a meeting he had not attended was an exceptional case (see the detailed reply to paragraph 47).

The Tacis case was an exceptional one.

Box 3

Phare

47. In this case the Polish contracting authority admits the absence of the evaluator at the last meeting of the evaluation committee. This evaluator sent the chairman of the committee a letter explaining his absence. The signing of the list of participants was probably the result of a mistake as to the date. In any event, absent or not, the marks of this evaluator in no way affected or nullified the final outcome of the invitation to tender.

48. All the evaluators signed the evaluation report, thus affirming the validity of the procedure (Bulgarian contracting authority).

49. In response to the Court's comments, the Commission delegation in Romania requested an explanation from the chairman of the local contracting authority's evaluation committee. He explained that the remarks of the evaluator quoted by the Court arose from the fact that some evaluators finished their individual evaluations more quickly than others and that all were present when the interviews were held. All members were also present at the final phase of the evaluation, namely the discussions and the reaching of conclusions concerning the committee's final recommendation. Furthermore, all the evaluators filled in their sheets and evaluation grids correctly, and the procedure was conducted in normal conditions of confidentiality in the room where the tenders were kept permanently under lock and key.

Marks awarded to the technical proposals

50. To ensure a good cost-benefit ratio and maintain the quality of projects costing millions of euro, it is better to avoid false economies when it comes to technical assistance. The fact that the best technical bids were chosen in almost all cases is fully consonant with the objectives of the Phare and Tacis programmes.

51. For both Phare and Tacis, the range of marks can vary markedly as a result of different mentalities and cultures. Evaluators work quite independently, and any significant divergence has to be justified. A procedure can only be cancelled if the marking reveals irregularities, not because of a wide range of marks, which can be completely justified and accepted by the evaluation committee and the contracting authority.

Box 4

Phare

52. In this case, (the Polish contracting authority), while it is true that the marks awarded by one of the evaluators diverged from those awarded by the other members of the committee, this same person awarded top marks to tenderers that did not win the contract. Thus any suspicion of bias is unfounded.

Tacis

53. In this case the results of the financial evaluation had most weight and led to the choice of the successful tenderer. The technical marks of the recipient's representative therefore did influence the award of the contract.

54. The Court's suggestion that the top and bottom marks be discarded is very difficult to put into practice since it would mean systematically summoning five members of an evaluation committee only to rule out the marks of two of them. Moreover, the chairmen of evaluation committees check whether any significant variation in the marks is the result of an evaluator's bias. If that turns out to be the case, either the evaluation is cancelled and referred to another committee or the marks of the evaluator in question are not taken into account.

Interviews

55. Both for Phare and Tacis, setting a precise percentage for the interview does not necessarily offer any methodological advantage. Indeed, interviews take place when the evaluators have examined the written technical proposals. The aim of interviews is to go into some aspects of the proposal in greater detail and get a better idea of the abilities of the team proposed. As soon as they have finished, the evaluators indicate whether or not and to what extent the interviews are likely to affect, favourably or unfavourably, their initial opinion. If necessary, they then modify some of their marks accordingly.

It is quite normal for final technical marks to be reconsidered following the interviews when it appears that, despite the quality of written bid, the experts' skills or suitability fall short. The weight of interviews may vary from one procedure to another in relation to the project in question, the complexity of the work and the role of each expert.

56. The option of interviews is taken on the basis of the nature of the project and the profiles sought. The possibility of interview is always mentioned in the instructions to tenderers and is the decision of the evaluation committee.

The practice of not inviting firms whose written technical bid does not meet the minimum quality threshold is a procedure comparable to an examination in two parts, written then oral. It avoids unnecessary expense for tenderers who have no chance of being awarded the contract.

Financial evaluation methods

57. Under both Phare and Tacis there were two methods of evaluating financial bids during the reference period:

(a) the total price: method used for projects where the expected result was clearly defined in terms of an end product or a precise number of man-days of specific expertise;

(b) the total price divided by the number of experts: method used for projects where the expected result was defined in terms of objectives and it was up to the tenderer to specify the type and duration of the technical assistance deemed necessary to achieve these objectives.

The two methods could be combined, depending on whether reimbursable expenditure was taken into account, or the cost of equipment, and depending on the specific nature of the invitation to tender.

It is not possible or desirable to lay down once and for all every last detail of a method and render it immutable. On the contrary, by combining evaluation methods we can take account of the individual features of each case, and so make the best choice.

For example, the cost of some components of a contract might be fixed and laid down in the specifications (the cost of equipment) and so it would be useless to take account of it in the financial evaluation.

58 and 59. For a team made up of a given number and type of experts, a private firm would be taking a major risk if it proposed a higher proportion of European experts than necessary to satisfy the technical specifications, given the considerable difference between the cost of an EU expert and the cost of a local one (of three to one). The alleged advantage accorded to EU experts is thus debatable.

This method of financial evaluation has been abandoned, however, because of the political drawbacks and its complexity.

Changes in the methods of financial evaluation

60. In the case of Phare, the Commission acknowledges that methods other than those announced should not have been applied by the contracting authorities concerned. However, it should be noted that in all the cases in question, the financial evaluation method was decided before the opening of technical and financial bids, which excluded any discrimination or favouritism in the award of these contracts.

Box 5

Phare

61 and 62. Errors were indeed made in these two cases.

63. In the Tacis cases cited by the Court a financial criterion other than that adopted before the opening of the tenders was indeed applied, for duly justified reasons. In any event, the changes had no influence on the outcome of the three invitations to tender in question.

File management

64. The dispersal of documents is inherent in the decentralised and devolved nature of a programme such as Phare, where the recipients are countries applying for accession. The Commission admits that in some cases, as a result of local circumstances, the documentation was not kept in a uniform way. The fact that Phare is decentralised means that it is extremely difficult to ensure that presentation is uniform and that the local contracting authorities routinely keep complete files. However, in June 1999 the Commission gave the delegations detailed instructions on the keeping of files.

As to Tacis, the Commission notes that the management of files, though not irreproachable, is acceptable.

65. The Commission acknowledges that, for the period under consideration by the Court, there were problems with the file-keeping for Phare. Measures have been taken since then to tackle these problems.

While not all financial bids were put in the Tacis files at the time, the evaluation reports, including the detailed minutes on procedures endorsed by the members of the committees, were routinely included.

The experts' activity reports are documents that are placed in the file of the evaluator's contract, not in that of the dossier for the tender in which they took part.

67. Instructions on the archiving of Phare files were given on 3 June 1999.

68. The management of information on contracts by a private operator is rendered necessary by the lack of statutory staff. At the moment the Commission does not have any choice but to farm out this work.

However, the Commission has always made sure that the risk of conflict of interest is kept to a minimum. Since the beginning of the Tacis programme, the contractor of the procurement unit has never been in charge of managing calls for tenders in which its partners took part. Since 1997, the PU contract has carried a clause excluding the contractor and all who work for him from Tacis contracts and financing.

69. The remark concerning the excessive turnover of managers is fully supported by the Commission. In the absence of sufficient statutory staff at both head office and in the delegations, recourse to non-statutory staff was and is the only solution. The periodic rotation of this non-permanent staff, which takes place at the end of contracts entails a loss of institutional memory, which does nothing to help file-keeping.

OBSERVATIONS CONCERNING FRAMEWORK CONTRACTS

72. The tendering procedure for new framework contracts covering all external aid is over. The contracts entered into force in April this year. To bridge the gap, existing Phare framework contracts were extended until the end of March.

74. On the subject of Tacis, the Commission does not share the Court's opinion that setting a minimum number of experts was a dubious condition. The contracting authority should be able to check tenderers' mobilisation capacity in terms of quality and quantity. This is by no means nullified by the possibility open to the Commission and the recipient to propose experts, because both options were provided for in the framework contracts. The evaluation committee excluded tenders on the basis of objective, clear and transparent controls, following the selection criteria set out in the tender dossier.

No firm has complained about the procedures used nor about the results published in the Official Journal.

75. The Commission does not question the logic of the Court's proposal; however, in the spirit of the administrative reform of the Commission now under way, the approach envisaged is that of decentralisation of responsibility to the authorising officers of the Directorates-General in the framework of clearer rules and more rigorous audits.

CONCLUSIONS AND RECOMMENDATIONS

76. The Commission attaches great importance to compliance with contract-award procedures. Since 1997 many measures have been taken to improve the procedural framework of service contracts. Without denying that there were problems in the reference period, the Commission stresses that, of the 68 cases listed by the Court, 36 concern exclusively the incompleteness of the files. Of the 32 others, only five Phare cases had a real effect on the outcome of the tendering procedure.

Competition between tenderers

77. Setting up consortia was a deliberate policy at the time, dictated by operational considerations and, in the case of Tacis, it was one of the provisions of the Regulation in force.

Transparency of procedures

79. Phare: the possibility of interview is always mentioned and the financial evaluation methods are explained. Without denying that the specific problems observed in the sample examined by the Court did occur, the Commission stresses that overall the errors had very little influence on the final outcome of the invitations to tender (five of the 80 cases).

Tacis: the award criteria and mention of the possibility of interview were included in the instructions to tenderers attached to the tender dossier. It is true that financial bids of the tenderers not selected were not routinely filed in the reference period under consideration by the Court (1994 to 1998). Files are now complete. In any event, the Commission notes that the errors in question had no influence on the final outcome of the invitations to tender.

81 (a). The methods and criteria for technical and financial evaluation are now routinely included in the tender dossier.

81 (b). The DIS manual contains detailed instructions for evaluation committees which have to be followed by the national contracting authorities.

81 (c) and (d). Delegations received instructions on the filing of financial dossiers and contracts by summer 1999.

81 (e). This procedure is always followed.

Effectiveness of the procedures

82. The Commission fully agrees with the Court's remark about excessive staff turnover. It would like to have sufficient statutory staff to avoid this situation arising and provide more continuity, something which is impossible with precarious staff.

Responsibilities of the parties involved in the procedures

83. The problems cited were quite circumstantial. Absences of evaluators, in so far as they remain within reasonable limits, are inevitable and natural (illnesses and various other impediments) on a committee made up of five members from very different environments. The same holds for the differences in marking.

84. The recommendation about better controls on participants and eliminating the risk of confusion of roles has already been implemented in the context of the new procedures.

The systematic discarding of the top and bottom marks is impracticable.

The Court recommends that an agent of the Commission attend the evaluation committees convened by the local contracting authorities. Implementing this suggestion requires sufficient Commission delegation staff in the recipient countries, which is not the case currently.

General conclusion

85. Since the reference period of the audit, many measures to improve the award of contracts have been introduced or are in hand:

- a new instruction manual for Phare and Tacis, applicable from this year (2000), that is designed to harmonise the procedures,

- the drafting of instructions to improve the filing system in the delegations,

- the creation of an SCR website where advance information on contracts, contract notices and follow-up information is posted in order to improve transparency,

- the standardisation of tender dossiers and contracts with a view to a more rational documentation system.

86. The Commission fully supports the Court's comments concerning the need to employ statutory staff for the work involved in tendering procedures. However, in the current context of staff shortages, it is impossible to do so and still keep big programmes like Phare and Tacis operating normally.