31997Y0526(01)

SPECIAL REPORT No 5/97 on management of the Community cereals trade involving export refunds, special import arrangements and regional aid schemes together with the Commission's replies (Submitted pursuant to Article 188c (4) (2) of the EC Treaty) (97/C 159/01)

CONTENTS

Paragraph Page

1. INTRODUCTION . 1.1-1.3 3

2. FIXING THE REFUND RATE FOR CERTAIN PRIMARY PRODUCTS . 2.1-2.8 3

General . 2.1 3

Revision of coefficients . 2.2-2.7 3

Common wheat flour . 2.3-2.4 3

Durum wheatmeal . 2.5 4

Malt . 2.6-2.7 4

Summary of observations . 2.8 4

3. EXPORT REFUNDS ON MALT AND COMPOUND FEEDINGSTUFFS 3.1-3.8 4

Malt . 3.1-3.5 4

General . 3.1-3.2 4

Adjustment of the refund . 3.3 4

Controlling the adjustment scheme . 3.4-3.5 5

Compound feedingstuffs . 3.6-3.7 5

Summary of observations . 3.8 5

Paragraph Page

4. STANDARD OF PHYSICAL CHECKS ON EXPORTED CEREALS . 4.1-4.12 6

General . 4.1 6

Port of Vlissingen/Netherlands . 4.2-4.4 6

Ports of Hull and Ipswich/United Kingdom . 4.5 6

Ports of Rouen and Dunkerque/France . 4.6-4.7 6

Andalusia/Spain . 4.8-4.11 7

Summary of observations . 4.12 7

5. SCRUTINY OF THE COMMERCIAL DOCUMENTS OF UNDERTAKINGS . 5.1-5.5 7

General . 5.1 7

Coordination of scrutiny effort . 5.2 7

Specific difficulties in determining correct quantities for refund payment purposes . 5.3-5.4 8

Summary of observations . 5.5 8

6. SPECIAL IMPORT ARRANGEMENTS FOR MAIZE AND SORGHUM INTO SPAIN AND PORTUGAL . 6.1-6.10 8

General . 6.1-6.7 8

Direct purchases on the world market . 6.8-6.9 9

Summary of observations . 6.10 9

7. SUPPLY OF CEREALS UNDER THE POSEIMA SCHEME . 7.1-7.9 9

General . 7.1 9

Forecast supply balance . 7.2-7.3 10

Link between aid and refund rates . 7.4-7.5 10

Transmission of benefits to end-users . 7.6-7.7 10

Inter-island transport costs . 7.8 11

Summary of observations . 7.9 11

8. GENERAL CONCLUSION . 8.1 11

Table 1 . 12

Replies of the Commission . 13

1. INTRODUCTION

1.1. The main part of this Report is concerned with export refunds, which continues the Court's work in auditing the operation of the export refunds system in the major market sectors (see Special Reports Nos 2/90 (1), 2/92 (2) and 7/93 (3) and observations in the 1995 Annual Report). In the four budget years 1991 to 1994, export refunds totalling 11 043 Mio ECU were paid on cereals exported to third countries. The budget lines relating to export refunds on cereals (within heading B1-100) combine certain cereal grains and related products for expenditure purposes. Complete details of amounts paid on each cereal product were unavailable at the Commission. Table 1 (annexed), presents data indicating only the volume of such products for refund purposes in the cereals sector.

1.2. The Court examined the following aspects of the operation of the export refund system for cereals:

(a) setting the refund rates for cereal products of primary processing (e.g. flour, semolina, malt);

(b) operation of the refund system as applied to the export of specialized cereal products, i.e. malt and compound feedingstuffs;

(c) standard of physical checks on exports carried out under Council Regulation (EEC) No 386/90 of 12 February 1990 (4) at the time of export;

(d) standard of a posteriori controls carried out under Council Regulation (EEC) No 4045/89 (5) including specific control difficulties found in Member States.

1.3. In addition to refunds, the Report also covers two specific schemes relating to the trade in cereals:

(a) the imports of maize and sorghum into Spain and Portugal under the GATT Agreement; and

(b) the supply of cereals under the Poseima scheme to the Azores region.

2. FIXING THE REFUND RATE FOR CERTAIN PRIMARY PRODUCTS

General

2.1. The export of products of primary processing (especially common wheat flour, durum wheatmeal and malt) to third countries is of major commercial importance with corresponding significance for refund expenditure. The refund on these primary products is based on the quantity of basic cereals necessary to manufacture a certain quantity of that product. Therefore, a processing coefficient is applied to the quantity of the processed product. For example, a coefficient of 1,51 means that 1 000 kilograms of a certain processed product is deemed to be produced from 1 510 kilograms of the basic cereal. Since 1991 the refund coefficient is obtained by adjusting the processing coefficient to take account of the benefits to be gained from selling valuable by-products either on the Community market or on third-country markets. The higher the coefficient, the higher the rate of refund.

Revision of coefficients

2.2. In Special Report No 2/90 (paragraphs 2.65-2.66), the Court commented on the basic refund rates and on the practice of granting special refund rates, in order to match export subsidies from competing suppliers, including for example, the United States. These special refund rates deviated from those set in the relevant regulatory provisions. In its reply, the Commission indicated that it would amend the basic coefficients, which had stood unaltered since 1967, also to take account of the new processing yields.

Common wheat flour

2.3. The processing coefficient for the highest quality flour stood unchanged at 1,51 from 1967 to 1982, when it was adjusted to 1,40. This 1,40 coefficient continued in force until 1 October 1991 when, following a review of all processing coefficients, it was further reduced to 1,37. With regard to this reduction the Commission, following consideration of the estimated rates of yields for flour milling given by the Member States and of other factors, initially proposed:

(a) a processing coefficient of 1,38 for the highest quality flour;

(b) a reduction in this coefficient to 1,35 for refund purposes after taking account of the value of by-products;

(c) pro-rata changes in the refund coefficients for flour of lower quality.

2.4. However, the Commission changed the original proposal of 1,35 to 1,37, with corresponding adjustments for the other categories. This revised proposal was accepted by the Management Committee for Cereals (consisting of representatives of the Commission and Member States) and adopted in Commission Regulation (EEC) No 2849/91 (6), thus increasing the relevant refund rates by about 1,5 % and the corresponding budgetary cost by approximately 13 Mio ECU for the period October 1991-December 1994. The evidence for this change, as set out in the Commission's file, was difficult to follow.

Durum wheatmeal

2.5. The Commission decided to restructure the processing coefficients for durum wheatmeal and to base the new rates on average quality meal. Initially the Commission proposed to reduce the coefficients in the three new categories (depending on the ash content) to 1,45, 1,30 and 1,20. However, this proposal was also changed in favour of exporters to 1,50, 1,34, and 1,26 in the same Management Committee meeting (see 2.4) and adopted in Commission Regulation (EEC) No 2849/91, which involved an increased budgetary cost of approximately 20 Mio ECU for the period October 1991-December 1994. Again, the evidence provided, as set out in the Commission's file, was difficult to follow.

Malt

2.6. The original processing coefficient in force from 1967 was 1,33. The Commission planned to change this coefficient in September 1991 from 1,33 to 1,30, a reduction of only 2,3 % after a period of almost 25 years. The only evidence of the introduction of the revised malt coefficient from this time is a reference in the Commission's internal weekly report. Commission Regulation (EEC) No 468/92 (7), which gave legislative effect to this change, entered into force in March 1992 i.e., 5 months later than the changes introduced for flour and durum wheatmeal. The Commission was unable to produce the file dealing with the review of the coefficient and adoption of the new yield rates for malt.

2.7. Euromalt, the working committee of the EC malting industry, recommended that the Commission should maintain a coefficient of 1,33. However, information provided to the Court by representatives of the malt producing industry itself of several Member States shows that the achieved technical progress could have led to coefficients between 1,26 and 1,20, i.e. a reduction of 5,3 %-9,8 %. A reduction of three percentage points (application of 5,3 % rather than 2,3 %) over the three years 1992-1994 could have resulted in a 16 Mio ECU saving to the budget.

Summary of observations

2.8. In conclusion, the Court notes that:

(a) apart from the interim revision for flour in 1982, the flat-rate processing coefficients established in 1967 should have been reviewed and revised at far more frequent intervals than the 25 years that was allowed to elapse between 1967 and 1991;

(b) furthermore, and to preclude the possibility of exporters receiving a double refund or benefit, one for the primary product and a second one for by-products, the value of by-products should have been taken into account during the years prior to 1991 (see paragraph 2.1);

(c) the limited evidence made available by the Commission of the justification for changing its original proposals in favour of the revised and finally-adopted coefficients (in 1991 and 1992) is difficult to follow. The effect of these changes benefitted trade interests at an increased budgetary cost of approximately 49 Mio ECU (see paragraphs 2.4-2.5, 2.7).

3. EXPORT REFUNDS ON MALT AND COMPOUND FEEDINGSTUFFS

Malt

General

3.1. Malt exports almost entirely comprise unroasted barley varieties and are intended for production of beer or grain-based spirits. In the four years 1991 to 1994, the Commission paid export refunds amounting to 3 031 Mio ECU on exports of barley and malt made from barley, representing 27 % of the total refunds paid on cereals during that period.

3.2. The bulk of export licences for malt are issued in October and November each year and are valid until 30 September of the following year. This long period of validity was granted to take account of the characteristics of the product and the market. Refund rates are normally fixed in advance at the time of issue of the licences.

Adjustment of the refund

3.3. Commission Regulation (EEC) No 1680/78 of 17 July 1978 (8) provides for the adjustment of the export refund on malt made from barley in store at the end of the marketing year or on malt held in store at the end of a marketing year (i.e. end-June) and is exported or placed under export control during the first three months of the following marketing year, (i.e. July, August, September). The effect of this provision is that the refund is paid at the pre-fixed rate (usually pre-fixed by the malt producing industry the previous October) and adjusted by means of monthly increments to the appropriate rate up to the following June. The refund rate applicable from the start of the new marketing year (July) has always been lower than the prefixed and adjusted rate (by the monthly increments) of the preceding year due to the decrease in the threshold price fixed by the Commission.

Controlling the adjustment scheme

3.4. Article 3 of Commission Regulation (EEC) No 1680/78 stipulates that the eligible malt is that drawn from the stocks held at the end of June (i.e., from the old crop). The substitution of malt from the new crop when using pre-fixed licences from the previous year could lead to commercial advantages for the exporter when exporters are contractually bound to supply their customers with new crop malt.

(a) In France doubts regarding the interpretation of this Regulation as regards the identity of the exported malt had the effect of suspending the customs checks between July 1993 and February 1994. The uncertainty concerned whether malt from the new crop can be used in substitution for the malt declared in stock at the end of the previous marketing year (i.e., end-June). The 'Commission Interministérielle de Coordination des Contrôles` in Paris considered in February 1994 that management on a substitution basis could be allowed provided that operators remained within the limits of the stocks left over from the previous year, as entered in the accounts at 30 June. This being so, the French customs authorities were informed by the 'Commission Interministerielle de Coordination des Contrôles` that there was no ground for arguing that an irregularity had been committed when a refund at the increased rate applying to the old year was paid on malt and barley from the new year, provided that this was within the limits of the amounts pertaining to the previous year, as entered in the accounts at 30 June;

(b) in the United Kingdom, the source of the malt declared under this scheme is not verified in the a posteriori scrutiny of export transactions. In May 1994, irregularities were discovered involving an estimated 3,3 Mio ECU in pre-financed refunds for malt during 1991 to 1993. In June 1996, these were subject of a legal procedure on possible criminal activities but no recovery of any of the amounts unduly paid has yet been made;

(c) Germany upheld the principle that the malt stocks declared for export under the adjusted refund must originate from the stocks declared at the end of the marketing year. Therefore, during the clearance of accounts procedure for 1988, the strict application of status rules resulted in a recovery of 6,1 Mio ECU.

3.5. The Commission has informed the Court that malt from the new crop may not be substituted for that from the old crop (see also paragraphs 3.3.-3.4). Given this clear position, the control situation found in France and the United Kingdom is unsatisfactory. It is of particular concern that this unsatisfactory situation which can easily lead to a distortion of competition in Member States should have been allowed to develop for a period of 18 years. The Court calls on the Commission to take urgent steps to have the correctness of payments verified, and where appropriate to recover amounts unduly paid.

Compound feedingstuffs

3.6. The refund is based on the percentage of cereals contained in the product. The definition of 'cereal product content` in the refund nomenclature excludes residue from the milling industry, e.g. vegetable waste, vegetable residue and by-products of such processing. The Court examined transactions relating to refund payments of approximately 3,6 Mio ECU made to a German beneficiary during 1992 and 1993 for exports of an animal feeding compound derived principally from the milling of maize. Refunds were paid, at the highest rate, because of the high cereal content declared. Reports made by the German customs inspection service under Council Regulation (EEC) No 4045/89 pointed out verification problems regarding processing of compound feedingstuffs. There was a difference of opinion between the customs laboratories in Hamburg and Berlin on the refund classification of the separate components contained in the exported product. Because of these uncertainties, the Court concluded that there were doubts over the classification.

3.7. Owing to the complexity of the refund and customs tariff nomenclature and rules and the difficulties in applying these rules to the specific manufacturing operations, it is not possible to be assured on the correctness in this case of the refunds paid. These problems were brought to the attention of the Commission. In its response, the Commission confined itself to a statement based essentially on the formal results of analysis supporting the exporter's declaration but provided neither an opinion on the areas of doubt and difficulty raised by the Court nor a statement on the eligibility of the exported product to the rates on which the refunds were paid.

Summary of observations

3.8. In conclusion, the Court notes that:

(a) failure to verify eligibility of the exported malt to the special adjusted refunds were found in two Member States (see paragraphs 3.4-3.5);

(b) in the case of refunds paid to a certain beneficiary, the Court has been unable to obtain a clear statement from the Commission on the correctness of the refund rates applied to the animal feedingstuffs (see paragraphs 3.6-3.7).

4. STANDARD OF PHYSICAL CHECKS ON EXPORTED CEREALS

General

4.1. Council Regulation (EEC) No 386/90 and further implementing Regulation introduced provisions to strengthen the physical monitoring of the export of products receiving refunds. In view of the importance of physical checks, the Court's audit included an examination of their effectiveness at five selected seaports in the Netherlands, United Kingdom and France in 1993 and at prefinancing storage and export locations in Spain in 1994. The seaports and export locations were selected according to their importance in the cereals market.

Port of Vlissingen/Netherlands

4.2. A high proportion of bulk grain exports sourced in the United Kingdom and France had been exported via the port of Vlissingen (Netherlands) to Russia and other republics of the former Soviet Union. These consignments were placed under export control, not at the port of loading, but at Vlissingen en route to destination. During the short stay at Vlissingen the grain remained on the vessel.

4.3. Since the grain was already loaded the checks carried out by the Dutch customs consisted of drawing samples from the ships' holds and having the exported quantity calculated by means of a draft survey technique (specific technique to measure the quantity loaded on board of a vessel). The accuracy of this method depends on a wide variety of technical factors and provides only an estimate of the loaded quantity. The Court considered checks carried out at Vlissingen inadequate for the following reasons:

(a) the export goods were not accessible for proper inspection and the drawing of representative samples for analysis was severely restricted due to the inconvenient and short docking periods of the vessels;

(b) the draft survey provided only an estimation of the loaded weight;

(c) the practice is at variance with normal control procedures in accordance with which the grain placed under export control is required to be made available for customs inspection prior to loading of the vessel with the customs authorities having the opportunity to interpose checks at any time during loading.

4.4. In principle, the rules governing the physical checks to be performed on export transactions involving the payment of refunds should preclude the possibility of such goods being presented for export examination at a transit port when already loaded on the export vessel. This practice of performing checks on goods loaded on board of export vessels was already criticised in the Court's Special Report No 2/90 (paragraph 3.33 (m)).

Ports of Hull and Ipswich/United Kingdom

4.5. Two seaports in the United Kingdom were visited to examine controls relating to the shipment of selected bulk grain exports on which refunds had been paid. Delivery of the grain (wheat) to the export vessel and loading operations were typically spread over several days. The following weaknesses were noted:

(a) the availability of the goods at the time of acceptance of the export declaration was rarely verified in cases where the grain was declared as stored remote from the Customs export office. In the port of Hull, attention had not been given to examining the silo records even when the silo was situated close to the export office;

(b) there was no reconciliation between quantities loaded with the quantities recorded as delivered from the storage silo to the export vessel. For example, in Ipswich, there was inadequate evidence of a supervisory customs presence at the grain terminal or the weighbridges during loading operations. At this grain terminal, grain is circulated and routed through the weighbridge. Therefore, the risk existed that weight tickets could be produced and presented for export purposes for grain simply circulating within the silo;

(c) in general, more frequent in-depth physical checks regarding operations and activities affecting the sourcing, quantity and description of the grain loaded for export were needed. In addition, there was a need to verify, at the material time, the existence of the stocks taken under export control.

Ports of Rouen and Dunkerque/France

4.6. The grain silos in the port of Rouen constitute one of the largest groups of storage and warehousing facilities for grain exports in the Community. Large stocks of open-market, intervention and prefinanced cereals are held in these silos for direct loading into ocean-going vessels. It is particularly important to be able to distinguish open-market stocks from intervention stocks which normally attract no refund or a reduced refund.

4.7. However, in the case of both intervention and open-market stocks, the physical storage location was not specified below silo level. Each silo complex comprises a large number of 'cellules` each capable of holding 4 000 to 6 000 tonnes of grain. These 'cellules` are used interchangeably for the storage of open-market or intervention grain depending on storage demand and are, for stock-control purposes, the storage units used for locating particular stocks. Since the precise location of stocks of different status is not indicated at cellule level in the silo records, reliance is placed on the knowledge of local silo supervisors. This situation is inadequate to guard against the risk of substitution of grain for refund purposes by intervention grain of similar or lower quality (see Commission Regulation (EEC) No 1776/92 of 30 June 1992). The same general situation was found in the Dunkerque silos.

Andalusia/Spain

4.8. For a representative beneficiary of the prefinancing scheme, the checks carried out in the Andalusia region were examined. Essentially, they comprised of two phases:

(a) designating the customs office responsible; and

(b) arranging inspection visits to various premises to verify the existence of the declared stock.

4.9. The delays involved in putting these arrangements into practice hampered the timely executions of the inspection visits, considering the short interval between notification of the beneficiary's intention to place the cereals under control and the date of commencement of the prefinancing procedure. As a result, the visits were too late for verification by the local authorities of the existence of the declared stock on the date of lodging the prefinancing declaration.

4.10. This beneficiary obtained an advance-fixing certificate for the exportation of 29 000 tonnes of durum wheat meal at the special refund rate of 210 ECU/tonne. This rate, set in September 1992, was about 70 ECU/tonne higher than the average refund rate for this product during the year 1993. The beneficiary in question 'advance-fixed` the above rate on 4 September 1992 and proceeded to place 25 637 tonnes under prefinancing control (9) on the last day of validity of the advance-fixing certificate. Although this was not against the Regulation, in these circumstances these transactions should have been selected for verification under risk-assessment criteria. However, as with all other 1993 prefinanced declarations lodged by this beneficiary, no physical checks were performed on prefinanced stocks prior to exportation and only one physical check was carried out in 1992 and in 1994. Considering the refund amounts at stake (about 5,4 Mio ECU) the degree of supervision was inadequate.

4.11. In October 1995, the Court wrote to the Spanish authorities regarding the following findings resulting from a visit to the same exporter:

(a) a discrepancy of roughly 385 tonnes of durum wheat between the quantity declared for prefinancing and the quantity in the exporter's commercial stock records on the same date;

(b) the substitution of over 5 000 tonnes of durum and common wheat from the open-market for stock previously placed under the prefinancing regime without permission of the competent authorities as required.

No response has, as yet, been received from the Spanish national authorities.

Summary of observations

4.12. A selective examination of the physical checks performed at important export locations revealed serious weaknesses in the organization of controls and in the methods used to verify the essential status and quantity of the goods exported (see paragraphs 4.1-4.11).

5. SCRUTINY OF THE COMMERCIAL DOCUMENTS OF UNDERTAKINGS

General

5.1. In its Special Report No 7/93 on the controls in the context of fraud and irregularities in the agricultural sector the Court criticized the implementation of Regulation (EEC) No 4045/89 by the Member States. In the course of the present audit it was found that a number of weaknesses, mentioned in the Special Report, continue to exist.

Coordination of scrutiny effort

5.2. The lack of coordination of the scrutinies carried out in the headquarters and those in the subsidiary companies in other Member States weakens the effectiveness of the exercise. The scrutinies of the subsidiary companies are necessarily limited in scope because they will not be able to cover all aspects of inter-company transactions. In cases involving complex dealings within a multinational corporate structure, scrutinies should be planned and executed on a joint basis by the national agencies concerned using appropriate means of coordination, such as that envisaged in Article 7 of Regulation (EEC) No 4045/89, with regard to exchange of information. However, there has been very limited use made by the scrutiny agencies concerned (France, Netherlands and United Kingdom) of this Article.

Specific difficulties in determining correct quantities for refund payment purposes

5.3. In the examination of refunds made to a major beneficiary based in the Netherlands for grain exports to the former Soviet Union, a regular element was the issue of credits to the buyer for weight shortages claimed to have been found at the port of unloading. These differences amounted to between 0,5 and 1,0 % of the invoiced and refund-declared quantities. While the shortage per ship was relatively small, almost every shipment eventually involved a credit claim. All claims were paid in full by the exporter without dispute. The exporter had decided for reasons of client goodwill to take this position and not because shortages on delivery had actually occurred. However, no adjustment of the export refund - which is paid on a weight basis - had been made for these documented shortages. This feature of cereal exports to the former Soviet Union territories was encountered not only in the Netherlands but also in Belgium, Germany and France.

5.4. The Member States' authorities brought the problem to the attention of the Commission. In its response, dated 7 April 1993, the Commission advised that, where differences in weight arise, the refund is due only on that quantity reaching the market of destination unless the differences can be ascribed to natural causes. However, the real issue concerned the position in relation to the recovery of refunds in the light of the credits issued by the exporters for weight shortages. Regrettably, clear guidelines on this issue were not given and no further action was taken. The Commission should have taken more decisive action in defining the effect of these credits on refunds and ensuring that a correct and uniform approach towards the recovery of any refunds due was adopted in the Member States.

Summary of observations

5.5. In summary, the Court observes that:

(a) there is a need for a joint audit approach by Member States in dealing with multi-national beneficiaries (see paragraph 5.2);

(b) the subject of credits in respect of weight shortages, had not been satisfactorily resolved or pursued (see paragraphs 5.3-5.4).

6. SPECIAL IMPORT ARRANGEMENTS FOR MAIZE AND SORGHUM INTO SPAIN AND PORTUGAL

General

6.1. To find a solution to the trade dispute existing between the Community and the United States, following the enlargement of the Community to include Spain and Portugal, Council Decision 87/224/EEC of 30 January 1987 approved an Agreement between the European Economic Community and the United States of America on the conclusion of negotiations under GATT Article XXIV.6. Under this Agreement, the Community undertook to ensure importation of a minimum of 2 Mio tonnes of maize and 300 000 tonnes of sorghum annually into Spain. This commitment was for the period 1987 to 1990 but has been extended as it was considered as appropriate on a year-by-year basis since 1990. Provision is made for reduction of these global quotas by the quantities of certain cereal substitutes imported during the year. The cereals may be purchased directly on the world market by the intervention agency or imported by commercial operators with a reduced import levy calculated on a flat-rate or tendering basis. Almost all imports have been carried out by commercial operators using the tendering system.

6.2. In the same context, the Community also agreed (Council Decision 93/355/EEC) to open an annual quota of 500 000 tonnes of maize for importation into Portugal under similar terms commencing with the 1993/94 marketing year. This scheme does not provide for an offset for imported cereal substitutes.

6.3. A fundamental condition in the operation of these special arrangements is that imports should not disturb the domestic or Community cereals markets. It is the Commission which proposes the maximum levy reduction to be adopted by the Management Committee (see 2.4) at a level which is, in any event, above the intervention price. The Court attempted to examine how the maximum levy reduction had been determined for selected adjudications where the reduction finally granted was appreciably higher than that initially calculated by the Commission. The records indicating the criteria used by the Management Committee were called for but could not be produced. Therefore, the audit of the criteria taken into account by the Committee has not been possible.

6.4. Responsibility for central administration of the scheme lies with the Commission which attempts to meet the target quota at the end of each year. The Commission has experienced practical difficulties in administering the Spanish scheme because of the inaccuracy of the import data for maize and sorghum imports supplied by the Spanish authorities. These difficulties were a consequence of the fact that the quantities for which tenders are opened are based on current estimates of the outstanding balance remaining to be imported. Serious discrepancies existed between the import figures for these cereals for the years 1992 to 1994 presented to the Court by the Spanish authorities and those previously transmitted to and acted upon by the Commission. Furthermore, delays had arisen in making the data available. For example, the data for the period January to April 1993 were not sent to the Commission until 1 October 1993 and those for July to September 1993 not until 3 February 1994.

6.5. Another area of difficulty concerns the offset from the global commitment of certain cereal substitutes (namely, maize gluten, brewers' grain and citrus pulp). These products represented roughly 30 % in 1993 and 40 % in 1994 of the grain quota. As with the import data supplied for the maize and sorghum, data collected and notified for the substitutes were found to be subject to error and revisions.

6.6. The present system, based on the use of monthly import data is prone to error and inaccuracies. The fact that, owing to the time-lags involved in correcting import data, the definitive quantities may not be available from the national authorities until after the particular annual exercise has been closed is a serious flaw in the system.

6.7. A more reliable method would be that based on the attributed quantities entered on the special import licences used for each specific tendering operation. These quantities represent actual imports and are verified by customs. The number of licences are manageable since only about 10 importing companies are normally involved in this trade. Furthermore, the quantities obtained from this source can be linked via the import licences to the specific tender opened by the Commission. For data on imported cereal substitutes, use of the trade statistics data would continue to be required due to the absence of a licensing system for these products. Overall, however, the method suggested would improve the reliability and timeliness of the data.

Direct purchases of the world market

6.8. A departure from the practice, as applied in Spain, that the maximum levy reduction is set by the Commission was the compromise concluded in the Council in December 1993 with Portugal (10) whereby the 250 000 tonnes of maize purchased directly by INGA, the Portuguese intervention agency, on the world market were permitted to be sold on the domestic market at the intervention price. In the years 1994 to 1995, INGA purchased half of the Portuguese import quota of maize which was in total 500 000 tonnes, directly on the world market. These purchases were treated as intervention measures and the resulting expenditure is met from Community funds. These imports are sold on the Portuguese market at the ruling intervention price which is substantially lower than the open-market price, while the reduced import levy is paid by INGA. This confers benefits on the Portuguese animal-feed processors and other commercial purchasers (however, implying disadvantages for maize producers) by transferring the incidence of the levy payment from the importers to the Community budget and selling the maize at the intervention price. This practice permits third-country maize to be placed on the Portuguese market at the intervention price thereby conferring a price advantage on such imports over similar Community goods.

6.9. An assessment of the costs for Community funding for the year 1994 has been carried out. The average purchase cost of 119,8 ECU/tonne and the average receipts from sales of 82,4 ECU/tonne resulted in operational losses of 37,4 ECU/tonne. In terms of the annual quota of 250 000 tonnes, this amounts to approximately 9,3 Mio ECU, levy included. This operational deficit stems primarily from the combined effect of INGA bearing the cost of the import levy and the sale of the maize at the low intervention price level.

Summary of observations

6.10. In conclusion, the Court notes that:

(a) owing to the non-availability of essential records, the Court has been unable to audit the criteria used in determining the maximum reduction in the levy to be granted (see paragraph 6.3);

(b) inaccuracies and delays in the import data supplied to the Commission for the purpose of administering this GATT Agreement has created severe difficulties in managing the scheme (see paragraphs 6.4-6.7);

(c) imported maize, purchased directly on the world market but sold at the intervention price on the Portuguese market, resulted in extra budgetary costs and is at variance with sound financial management principles (see paragraphs 6.8-6.9).

7. SUPPLY OF CEREALS UNDER THE POSEIMA SCHEME

General

7.1. The POSEI (Programme of Options Specifically Relating to Remoteness and Insularity) aid programmes are designed to remedy, as regards the supply of specified agricultural products, the exceptional geographical situation of certain remote, insular regions (11).

Forecast supply balance

7.2. In accordance with the basic Council Regulation (EEC) No 1600/92 (12), a forecast supply balance for the agricultural products necessary for consumption and processing is determined on an annual basis for each eligible product. Within this quantitative limit, these products may be imported into the region concerned without import levies (products of third-country origin) or with an aid amount giving an equivalent price advantage (Community products).

7.3. The cereals supply balance for the Azores for the first year of operation (1992/93) amounted to 150 000 tonnes - this quantity was subsequently agreed for years 1993/94 and 1994/95. However, the Commission could not provide the documentation showing how the original balance was arrived at and the approval for its adoption.

Link between aid and refund rates

7.4. Almost the entire cereals supply balance shipped to the Azores has been sourced in the Community - (mainly, French maize for animal feeding). The level of aid normally equates with the rate of export refunds in force for the same products at the time of issue of the aid certificates. This mechanism has resulted in:

(a) variability in aid rates from month to month;

(b) a steady decrease in the rates from 1992 to 1994 and a more dramatic fall in marketing year 1995/96 due to the decline in refund rates for cereals (13); with the refund linkage, it is inevitable that factors irrelevant to the specific requirements of the POSEI scheme will dictate aid levels.

This instability in aid rates cannot serve the objectives of the Poseima supply arrangements which seek to remedy certain structural, long-term disadvantages of the region. More generally, the use of a product-based subsidy has been inappropriate and has led to the inefficient use of Union funds.

7.5 Moreover, the Court found that a discrepancy existed between the aid rate set for maize in October 1993 (105 ECU/tonne) and the export refund in force during the same month (35 ECU/tonne). The aid rate was not reduced in line with the refund rate until one month later (to 43 ECU/tonne). The Commission explained that the aid rate for October 1993 had been higher than strictly necessary in the circumstances. In fact the October rate should have been approximately equal to the November rate of 43 ECU/tonne and was, therefore 62 ECU/tonne too high. The quantity imported under October aid certificates was 42 660 tonnes for Poseima and 34 427 tonnes for the Poseidom scheme (French overseas departments). The cost of this excessive aid rate to the Community budget is estimated at 2,6 Mio ECU for Poseima and 2,1 Mio ECU for Poseidom. A better method of providing a remedy for the supply disadvantages of the POSEI regions might be an annual subsidy calculated to compensate for the additional transport costs incurred in conducting trade with the mainland Union territory.

Transmission of benefits to end-users

7.6. The transmission of the aid benefits to the end-user is an important requirement of the supply arrangements. In common with all POSEI schemes, the competent national authorities are required to take all necessary measures to ensure that the aid benefits are passed on and to recover benefits where this has not been done. However, the extent to which the benefits have been passed on is not clear from the accounts of the importing and processing companies visited. Whereas the aid amount is deducted in calculating the selling price, this deduction can be recovered in the profit margins.

7.7. In 21 selected transactions representing sales of imported grain carried out by the principal Azorean importer between 1992 and 1994, the percentage mark-up (i.e. the percentage difference between the net import cost - after deduction of aid - and the invoiced sale price), ranged between 12,1 and 41,7 % for sales to a major customer who handled his own storage costs and between 29,3 and 81,8 % for other clients. The wide range in these percentages are partly explained by the variation in the aid amount. At times where the aid rate is relatively high in relation to the purchase price, the first seller in the Azores has the opportunity to take a higher profit while keeping his selling price unchanged. In general, therefore, the transmission of the aid benefit is not transparent in the sale of cereals by importers and these benefits are only partially passed on in the selling prices. Considering the dominant position of this principal importer, there was a notable absence of active scrutiny and control by the regional authorities.

Inter-island transport costs

7.8. The present aid scheme does not take into account inter-island transport costs in the distribution of cereals to the smaller islands in the Azorean archipelago that lack the port facilities to handle direct bulk imports. While provision was made in Article 1 of Commission Regulation (EEC) No 1727/92 (14), for the allocation of the cereal supplies to the individual islands within the Azores, it was found that local problems had arisen relating to inter-island distribution services and delivery charges. It would clearly be more satisfactory to have an aid scheme that incorporates an element designed to cover inter-island transport costs.

Summary of observations

7.9. The Court notes that:

(a) the use of export refunds as the basis of determining aid rates has resulted in wide fluctuations in aid payments and, since 1995, to the virtual disappearance of aid on cereals (see paragraphs 7.4-7.5);

(b) the controls aimed at ensuring that the price benefits are transmitted to the end-user were inadequate and permitted high trading margins to be taken by certain cereal importers (see paragraphs 7.6-7.7).

(c) the use of product-based subsidies do not appear to be a sound basis for granting relief to compensate for the geographical and other disadvantages of the Poseima regions - in particular, the specific aspect of inter-island transport costs to the smaller Azorean islands lacking adequate port facilities (see paragraph 7.8).

8. GENERAL CONCLUSION

8.1. On the basis of the audit findings, the Court concludes that:

(a) failure of the Commission to revise the processing coefficients, forming the basis of the refund rates paid on primary cereal products, for a period of 25 years has led to the payment of refunds higher than that justified by actual production yields (see paragraphs 2.1-2.2);

(b) difficulties had occurred in finalizing the audit of certain aspects of trade in cereals partly due to the non-availability of documentation and partly due to decision-making procedures which were hard to follow (see paragraphs 2.4-2.5, 2.7 and 6.3);

(c) there are serious control weaknesses in the verification of the adjusted refund payments made under the special scheme for malt (see paragraphs 3.4-3.5);

(d) the rules for paying refunds on compound feedingstuffs have not been sufficiently well designed to ensure that refunds are paid only on the eligible cereal content (see paragraphs 3.6-3.7);

(e) the organization and execution of physical checks by the Member States need to be improved to ensure the correctness of the status, description and quantity of exported cereals (see paragraphs 4.1-4.11);

(f) the scrutiny of multinational beneficiaries should be carried out on a joint audit basis (see paragraph 5.2);

(g) in implementing the special import Agreement for maize and sorghum into Spain, the quality of the import data, forming the basis for administering the commitment, needs to be radically improved (see paragraphs 6.4-6.7);

(h) the extra budgetary costs involved in the sale of imported maize at the intervention price on the Portuguese market cannot be justified (see paragraphs 6.8-6.9);

(i) the supply of cereals under the Poseima aid scheme suffers from a number of serious defects, notably linking the aid rates to the export refund rates (also for other POSEI schemes). A remedy for the supply disadvantages of the POSEI regions might be an annual subsidy calculated to compensate for the additional transport costs between the regions and the mainland Union territory (see paragraphs 7.4-7.8).

This report was adopted by the Court of Auditors in Luxembourg at its meeting of 10 April 1997.

For the Court of Auditors

Bernhard FRIEDMANN

President

TABLE 1

>TABLE>

(1) OJ No C 133, 31. 5. 1990.

(2) OJ No C 101, 22. 4. 1992.

(3) OJ No C 53, 19. 2. 1994.

(4) OJ No L 42, 16. 2. 1990.

(5) OJ No L 388, 30. 12. 1989.

(6) OJ No L 272, 28. 9. 1991.

(7) OJ No L 53, 28. 2. 1992.

(8) OJ No L 193, 18. 7. 1978.

(9) Subject to financial guarantees, a facility for the advance payment of the refund exists whereby the refund is paid prior to export at the time of placing the products (or basic materials) intended for export under customs control for storage or processing purposes.

(10) OJ No L 83, 26. 3. 1994, p. 26.

(11) In total there are three POSEI aid programmes for ultra-peripheral areas, which are:

- Poseican: the programme for the Spanish Canarian Islands;

- Poseidom: the programme for the French Overseas Departments;

- Poseima: the programme for the Portuguese Azores and Madeira.

(12) OJ No L 173, 27. 6. 1992.

(13) For maize from about 100 ECU/tonne in 1992/93 to about 70 ECU/tonne in 1994/95 and for common wheat from about 73 ECU/tonne in 1992/93 to about 44 ECU/tonne in 1994/95.

(14) OJ No L 179, 1. 7. 1992.

REPLIES OF THE COMMISSION

1. INTRODUCTION

1.1 Details of the amounts paid for export refunds in respect of each combined nomenclature code for cereal products regarding exports from the principal exporting countries are available at the Commission. They are not, however, complete, as some Member States' Paying Agencies do not record these details.

2. FIXING THE REFUND RATE FOR CERTAIN PRIMARY PRODUCTS

2.1 and 2.8 (b) The coefficients set by Regulation (EEC) No 2849/91 are a marked improvement on those in Regulation (EEC) No 162/67. They take account of a standard value for by-products (e.g. bran). The value derived from by-products is extremely variable over time and between Member States and to try to take it into account in a precise way would mean doing the same for differences in the price of the product to be processed between Member States and differences in it dictated by the geographical location of each processing plant reflecting forward and internal transport costs etc. Such an approach would be extremely difficult to operate and would certainly lead to distortion of competition and higher budget costs as it would be impossible from Brussels to monitor the accuracy of all the data required.

A standard value approach is in fact the only way of guaranteeing transparency to users of the refund.

As regards value derived from by-products it should be pointed out that there is not necessarily a refund on all of these and that in the case of for example common wheat flour the slight advantage obtained by Community operators has been reflected in a corresponding fall in the processing coefficient from 1,39 to 1,37. Depending on international market prices this advantage could one day become a disadvantage.

Revision of coefficients

2.4 to 2.6 The processing coefficients adopted on each occasion are the outcome of consultation in the Management Committee for Cereals on the basis of the best technical information available to the Commission and each Member State at the time. This is why the coefficients of the Commission's original proposal of 4 July 1991 were revised slightly upwards.

The differences between the Commission's proposals were the outcome of research in each Member State. The first proposal, of January 1991, was to reduce the coefficients given in Regulation (EEC) No 162/67. Its sole purpose was to obtain the reactions of the Member States and industrial processors with a view to achieving the best technico-economic picture.

Industrial processors in the Community vary very widely in efficiency (installations of varying degrees of modernity) and size, and vary too in capacity, the outcome being widely differing economies of scale. The quality of the raw material treated can vary widely from one year or one region to another and thus technical yields can be very variable. This is particularly so for malt, which is produced from widely differing barleys.

Common wheat flour and durum wheat meal

2.4 to 2.5 It will be seen in the working document of 11 July 1991 on the proposal for a coefficient of 1,37 for common wheat flour that the coefficients notified by the Member States range widely from 1,32 to 1,43. The biggest flour exporters (France, Italy, Spain, Belgium, Netherlands), which have the most efficient processors, all give coefficients above 1,37 and their reasons for so doing led the Commission to alter its initial position.

The USA, a serious competitor of the EU on the world market, uses the same coefficient for flour (1,37) as the Community and for durum wheat meal a markedly higher one (1,723).

Malt

2.6 The five-month gap between publication had no financial implications. The new coefficients (1,3 for barley malt and 1,52 for roasted malt) were used with effect from 1991/92 marketing year.

The Commission has no file on this matter. The only document covering review of the malt coefficient is indeed the weekly report 'Cereals` No 6 (week of 9 to 13 September 1991).

2.7 The Commission was officially informed by Euromalt in July 1991 that the 1,33 coefficient should be reconfirmed. Most Member States firmly supported Euromalt's proposal.

Some well localized modern maltings may indeed have higher yields but how representative are they and on what scale are they working? It would be useful for the Commission to be sent any verified information the Court has.

Summary of observations

2.8 (a) and (b) The processing coefficients established in 1967 were essentially for technical guidance. Developments in the Community processing industry since 1967 as well as increased demand for animal feed (and hence for by-products) led the Commission to undertake a review of the processing coefficients. In its answer to the Court's observations on the review of processing coefficients for common wheat flour, durum wheatmeal and malt, the Commission has sought to give the background to the complex issues involved in undertaking such a review.

A single flat-rate coefficient is applied to each product because a differentiation of coefficients between each processor and product, of highly variable natures, would be extremely onerous to manage and the cost of controls would outweigh the benefits. A single flat-rate is inevitably only an approximation of the average production yield and cannot be fine-tuned on a frequent basis.

(c) The Commission's first proposal was adapted in the light of the evidence supplied by the Member States in the Management Committee, to a figure on which a consensus of technical experts could be obtained. The Commission has provided the Court with all the evidence at its disposal.

3. EXPORT REFUNDS ON MALT AND COMPOUND FEEDINGSTUFFS

Controlling the adjustment scheme

3.4 to 3.5 The Commission confirms that granting the adjustment of the export refund (Regulation (EEC) No 1680/78) for barley to be processed into malt or for malt in storage and to be carried over to the new crop year is limited to products of the old crop year. No substitution with produce coming from the new crop year is possible except when the provision regarding equivalence under prefinancing apply, in which case, however, the rate of the new crop year must be applied.

Compound feedingstuffs

3.6 to 3.7 The Commission confirms that:

(a) the export refund on compound feedingstuffs is granted on the actual cereal content. The products to be considered as cereals are clearly defined in the footnote appearing in

- the general Regulation (EEC) No 3846/87 last amended by Regulation (EC) No 2230/96) establishing an agricultural product nomenclature for export refunds, and

- the monthly Regulations setting the refunds on cereal-based compound feedingstuffs;

(b) definition of the cereal products in question is regulated by the explanatory notes to the Harmonized Commodity Description and Coding System.

The Court refers at point 3.6 to a specific case of refund payments on a mixture of co-products containing brans of heading 2302 of the Tariff Code. The Commission considers that this mixture cannot be covered by the code headings for products eligible for refunds: the cereal product in question for which the refund is sought belongs to heading 1104. The explanatory note for this heading excludes the presence of residues covered by heading 2302.

There appear to be some difficulties in the interpretation of the combined nomenclature, and the Commission will look into this question in more detail.

Summary of observations

3.8 Regarding the Court's conclusions

(a) financial consequences of any failure to verify the export refund claims in question which is demonstrated by the Court's documentation will be drawn in the clearance of accounts procedure;

(b) the Commission considers that the regulations are explicit as to the nature of products which are eligible for refunds. It will, however, clarify with the Member States any doubts they raise on interpretation.

4. STANDARD OF PHYSICAL CHECKS ON EXPORTED CEREALS

4.2 to 4.11 The Court's findings have been noted with interest. They confirm those of the Commission, established in their current control programme, under the clearance of accounts of the EAGGF, and which focus on the implementation of Regulation (EEC) No 386/90 (physical control) and on the management and control, by customs, of the prefinancing schemes (storage as well as processing).

In the light of these findings, initiatives are being taken to strengthen the legislative conditions relating to the place where the export declaration has to be lodged. Discussions on draft legislation, codifying the horizontal regulation on export refunds Regulation (EEC) No 3665/87), are currently taking place in the experts group on trade mechanisms.

The Commission has proposed financial consequences for the Netherlands (4.2 to 4.4) and France (4.6 to 4.7) for prefinancing of cereals for both EAGGF years 1993 and 1994. The prefinancing enquiry is currently being continued in the United Kingdom (4.5) and Spain (4.8 to 4.10).

5. SCRUTINY OF THE COMMERCIAL DOCUMENTS OF UNDERTAKINGS

Coordination of scrutiny effort

5.2 The Commission fully shares the Court's opinion on the need for coordination of scrutinies of international companies, through mutual assistance among the Member States' scrutiny services (within the same Member State as well as between Member States) in order to carry out controls down and up stream, particularly regarding international traders operating without any territorial limitations. This is certainly true for the export of cereals.

In all its control visits to Member States and in meetings of the working group on ex-post audits, the Commission insists on the need to intensify the use of mutual assistance. Concrete arrangements have been proposed to all Member States and the Commission will assess their impact.

Specific difficulties in determining correct quantities for refund payment purposes

5.3 to 5.4 With regard to the Court's observations on bulk cereals exports to the Republics of the former Soviet Union, and in particular the question of credit notes, the Commission, in its letter of 7 April 1993, considers that it gave clear guidance to the Member States. Member States were also requested to pay particular attention to this aspect of the cereals trade in the context of their scrutinies under Regulation (EEC) No 4045/89. Under the clearance of accounts procedure, the Commission requested controls on this area and is currently gathering the results of the controls which Member States have carried out as it requested.

5.5 In response to the Court's observations:

(a) The Commission agrees with the Court's observations on the need for a joint audit approach by Member States in dealing with multiannual beneficiaries. Work is being undertaken in this area in the working group on Regulation (EEC) No 4045/89.

(b) The Commission considers that it gave clear instructions in its letter of 7 April 1993. The Commission is currently gathering results of the controls undertaken by the Member States in this area.

6. SPECIAL IMPORT ARRANGEMENTS FOR MAIZE AND SORGHUM INTO SPAIN AND PORTUGAL

USA/EEC maize/sorghum agreement for Spain

6.3 to 6.7 The Commission's objectives in managing the quota are to respect Spanish market prices, and at a minimum the intervention price in production zones, and also intra-Community trade flows. It is moreover bound to honour its international import commitments. For the latter obligation to be respected maize and sorghum imports under the quota must be economically attractive to users in central Spain. For this to be so calculated prices for port zones, increased by the cost of transport to the interior of the country, must be competitive with market prices recorded in consumption zones.

The objective of importing without excessively unbalancing the single Community market is achieved with varying success from one marketing year and from one tendering procedure to another. The theoretical and mathematical calculations made in preparation for each weekly tendering procedure take as much account as possible of current market factors ('US` maize prices, exchange rates, freight, import duty, Spanish prices, transport costs within Spain, French prices for maize delivered in Spain etc.). The Commission however has to depend on the offers made by operators, which of course include a profit margin.

The international import commitment is for a minimum amount (2 million tonnes maize + 300 000 tonnes sorghum less cereal substitute products (CSP)). The agreement permits these quantities to be exceeded.

6.6 It is true that practical difficulties regularly occur in assessment of the CSP quantities imported. Here the Commission depends on information from the Spanish customs authorities. On-going close cooperation has now been established and the situation has markedly improved since the date of the Court's findings.

6.7 As the Court recommended, the Commission is now calculating the balance to be imported from the attributed quantities entered on the import licences issued for the relevant CSPs and maize and sorghum.

Direct purchases on the world market

6.8 to 6.9 The obligation to import maize into Portugal has so far been discharged in part through INGA. The Court remarks on the real budget cost of this arrangement introduced at the express request of Portugal approved by the Council. The Commission has been aware since it was first applied of the distortion involved. Taking its stand on the Court's remarks it will propose for 1997 that the arrangement be discontinued.

Summary of observations

6.10 In response to the Court's observations

(a) the grounds of the decisions taken by the Management Committee will in future be recorded and kept available to the Court;

(b) the flow of data from Spain has been improved;

(c) the arrangements for importing maize into Portugal criticized will not be proposed in 1997.

7. SUPPLY OF CEREALS UNDER THE POSEIMA SCHEME

7.3 The first supply balances (1992/93 marketing year) were set, in the course of negotiation of these programmes by the national authorities and the Commission, in line with local consumption requirements as presented by the relevant national authorities. Since then a constantly improved procedure has operated whereby the national authorities make reasoned applications that are examined by the Commission with due regard to what happened in the previous marketing year. The Commission is drawing up a report on the impact of the programme as part of the SEM 2000 exercise.

7.4, 7.5 and 7.9 The Commission's Interdepartmental Working Party for the outermost regions has begun an in-depth examination covering the method of calculating supply aid and use of the export refund as basis of calculation. The aim is to improve the effectiveness and stability of the arrangements.

Transmission of benefits to end-users

7.6 to 7.7 The Commission has several times raised with the national authorities the requirement to check that the benefit of the aid is passed on to the end-user. The actual definition of the final user varies by product and is not always sufficiently clear. The regional authorities have set up a control system to ensure that supply aids are properly reflected in consumer prices. This involves regular inspection of the various importers and retailers. These checks will be monitored by the Commission.

The Azorean regional authorities say that the supply scheme has helped increase the variety of operators carrying out inter-island cereal distribution.

Inter-island transport costs

7.8 The inter-island distribution factor will be studied by the Interdepartmental Working Party as part of its current examination. The Court's suggestion will be considered by the Commission for the programme's future.

Summary of observations

7.9 In response to the Court's observations:

(a) to (c) The Commission will discuss with the Portuguese authorities how the scheme can best be improved.

8. GENERAL CONCLUSION

8.1 (a) In its answer to the points raised in Chapter 2, the Commission has sought to highlight those issues which lay behind the review of the processing coefficients in 1991 and those which might lie behind future decisions to review these coefficients. However, it would stress the difficulties involved in varying the coefficients for individual exports, as this implies a coefficient for every single mill, and because of the variations in the quality of cereals according to regions and years. Such a system would also require a control burden not justified by any saving to the budget.

The Commission remains available to give the Court all explanations which it requires concerning the documentation supplied.

(c) The consequences of any control failings demonstrated by the information provided by the Court will be dealt with in the clearance of accounts procedure.

(d) In its answer to points 3.6 to 3.7, the Commission has sought to clarify the issues raised by the Court. The rules are explicit, and any doubts on their interpretation will be clarified with the Member States.

(e) The Commission has taken note of the Court's findings on the organization and implementation of physical checks in the Member States. Work is being undertaken on the codification of Regulation (EEC) No 3665/87 in the experts group on trade mechanisms. Control failings will be dealt with in the clearance of accounts procedure. Corrections have already been applied for failures to control the prefinancing regime.

(f) The Commission shares the Court's view on the need to undertake scrutinies of multinational beneficiaries on a joint audit basis. This matter is being discussed with the Member States in the working group on Regulation (EEC) No 4045/89.

(g) The Commission is cooperating closely with the Spanish authorities on the question of data transmission. Improvements have already been noted and further improvements may be expected in the context of the Customs 2000 initiative.

(h) The Commission shares the views of the Court and is considering undertaking a review during 1997 of the system currently in operation.

(i) Scrutiny of the various measures of the Poseima programme has shown that on the whole it has functioned satisfactorily. The Commission's aim will, in the light of an assessment exercise soon to be undertaken, be to tailor the programme more precisely to the needs of the two regions in question. The Court's suggestion for incorporation of inter-island transport costs will be considered.