(i) Article XIII:1

26. In this regard, we note that under the revised regime, on the one hand, bananas may be imported under the MFN tariff quota on the basis of past trade performance by exporting countries during a previous representative period (i.e. the three-year period from 1994 to 1996). On the other hand, bananas from traditional ACP supplier countries may be imported up to a collective amount of 857,700 tonnes, which was originally set to reflect the overall amount of the pre-1991 best-ever exports by individual traditional ACP suppliers, with allowance made for certain investments. ¹⁸⁶ We further note that exports under the tariff quota by some non-substantial suppliers (i.e. third-country and non-traditional ACP suppliers) are restricted, in aggregate, to 240,748 tonnes (i.e. the "other" category of the MFN tariff quota), whereas exports from other non-substantial sources of supply (i.e. traditional ACP suppliers) are restricted, in aggregate, to 857,700 tonnes. Moreover, some non-substantial suppliers, namely the ACP suppliers, could benefit from access to the "other" category of the MFN tariff quota once the 857,700 tonne tariff quota was exhausted. On the other hand, non-substantial suppliers from third countries have no access to the 857,700 tonne tariff quota once the "other" category of the MFN tariff quota is exhausted. Individual Members in these two groups � traditional ACP suppliers and the other non-substantial suppliers � are accordingly not similarly restricted. This disparate treatment is inconsistent with the provisions of Article XIII:1, which require that "[n]o � restriction shall be applied by any Member on the importation of any product of the territory of any other Member � unless the importation of the like product of all third countries � is similarly prohibited or restricted".

(ii) Article XIII:2

27. The general rule laid down in Article XIII:2 of GATT requires Members to "aim at a distribution of trade � approaching as closely as possible the shares which the various Members might be expected to obtain in the absence of such restrictions". To this end, where the option of allocating a tariff quota among supplying countries is chosen, Article XIII:2(d) provides that allocations of shares (i.e. country-specific allocations for substantial suppliers; and a global allotment in an "other" category for non-substantial suppliers unless country-specific allocations are allotted to each and every non-substantial supplier) should be based upon the proportions supplied during a previous representative period. The European Communities explains that it chose the three-year period from 1994 to 1996 as the most recent three-year period for which reliable import data were available.
28. According to the information available to us, for traditional ACP supplier countries the average exports during the three-year period from 1994 to 1996 were collectively at a level of approximately 685,000 tonnes, which is only about 80 per cent of the 857,700 tonnes reserved for traditional ACP imports under the previous as well as under the revised regime. In contrast, the MFN tariff quota of 2.2 million tonnes (autonomously increased by 353,000 tonnes) has been virtually filled since its creation (over 95 per cent) and there have been some out-of-quota imports. Thus, the allocation of an 857,700 tonne tariff quota for traditional banana imports from ACP States is inconsistent with the requirements of Article XIII:2(d) because the EC regime clearly does not aim at a distribution of trade approaching as closely as possible the shares which various Members might be expected to obtain in the absence of restrictions.
29. In light of the foregoing, and in light of the Appellate Body findings that the Lomé waiver does not cover inconsistencies with Article XIII, we find that imports from different non-substantial supplier countries are not similarly restricted in the meaning of Article XIII:1 of GATT. Moreover, we find that the allocation of a collective tariff quota for traditional ACP States does not approach as closely as possible the share which these countries might be expected to obtain in the absence of the restrictions as required by the chapeau to Article XIII:2 of GATT. Therefore, we find that the reservation of the quantity of 857,700 tonnes for traditional ACP imports under the revised regime is inconsistent with paragraphs 1 and 2 of Article XIII of GATT.

(c) The Requirements of the Appellate Body report in Bananas III

30. The European Communities recalls that the panel and the Appellate Body held in Bananas III that it is required by the Lomé Convention to provide duty-free access to traditional exports from ACP suppliers in an amount of their pre-1991 best-ever exports (i.e. 857,700 tonnes) and that the Appellate Body held that it could not assign country-specific allocations to those suppliers inconsistently with Article XIII. It argues that in consequence the Appellate Body report in Bananas III requires it to provide a collective allocation of 857,700 tonnes to those suppliers.
31. We note, however, that the panel and Appellate Body reports made it clear that what was required by the Lomé Convention was not necessarily covered by the Lomé waiver. And, as the Appellate Body found in Bananas III, the European Communities is not authorized by the Lomé waiver to act inconsistently with its obligations under Article XIII. The Appellate Body also upheld the panel finding that the European Communities could not allocate country-specific shares to some non-substantial suppliers (e.g. traditional and non-traditional ACP countries and BFA signatories) unless country-specific allocations were also given to all non-substantial suppliers.
32. We stress that the foregoing analysis does not render the Lomé waiver meaningless (see paragraphs sections D.4 and F below). We have taken appropriate account of the EC's admonition that we should not interpret Article XIII so as to reduce the Lomé waiver or Article I to inutility. Nor have we added to or reduced the rights or obligations of Members contrary to Article 3.2 of the DSU.

2. Ecuador's Share of the MFN Tariff Quota

33. Article XIII:2(d) provides that if a Member decides to allocate a tariff quota it may seek agreement on the allocation of shares in the quota with those Members having a substantial interest in supplying the product concerned. In the absence of such an agreement, the Member:

"shall allot to Members having a substantial interest in supplying the product shares based upon the proportions, supplied by such Members during a previous representative period, of the total quantity or value of imports of the product, due account being taken of any special factors which may have affected or may be affecting the trade in the product" (emphasis added).

34. Ecuador challenges the EC's allocation of the MFN tariff quota to it on the grounds that its share does not approximate the share that it might be expected to obtain in the absence of restrictions. It also argues that given the history of trade-distortive EC banana measures, it is far from clear that any country-share allocation system could be devised based on the idea of a representative period and special factors that would meet the requirements of Article XIII:2 (see paragraphs 6.47-6.48).
35. The European Communities notes that it based its calculation of country allocations under the MFN tariff quota of the revised regime on the three-year period from 1994 to 1996. In the EC's view, this was the most recent three-year period for which reliable data were available at the time.

(a) The Requirements of Article XIII

36. In considering Ecuador's claims regarding its tariff quota share under Article XIII, we recall our findings in Bananas III:

"The wording of Article XIII is clear. If quantitative restrictions are used (as an exception to the general ban on this use in Article XI), they are to be used in the least trade-distorting manner possible. In the terms of the general rule of the chapeau of Article XIII:2:

'In applying import restrictions to any product, Members shall aim at a distribution of trade in such product approaching as closely as possible the shares which the various Members might be expected to obtain in the absence of such restrictions �'

In this case we are concerned with tariff quotas, which are permitted under GATT rules, and not quantitative restrictions per se. However, Article XIII:5 makes it clear, and the parties agree, that Article XIII applies to the administration of tariff quotas. In light of the terms of Article XIII, it can be said that the object and purpose of Article XIII is to minimize the impact of a quota or tariff quota regime on trade flows by attempting to approximate under such measures the trade shares that would have occurred in the absence of the regime." ¹⁸⁷

37. We also noted the following:

"[I]n order to bring its banana import regulations into line with Article XIII, the European Communities would have to take account of Article XIII:1 and XIII:2(d). In order to allocate country-specific tariff quota shares consistently with the requirements of Article XIII, the European Communities would have to base such shares on an appropriate previous representative period³⁷⁵ and any special factors would have to be applied on a non-discriminatory basis."

______________________________
³⁷⁵ "In this regard, we note with approval the statement by the 1980 Chilean Apples panel:

'[I]n keeping with normal GATT practice the Panel considered it appropriate to use as a 'representative period' a three-year period previous to 1979, the year in which the EC measures were in effect. Due to the existence of restrictions in 1976, the Panel held that that year could not be considered as representative, and that the year immediately preceding 1976 should be used instead. The Panel thus chose the years 1975, 1977, 1978 as a 'representative period'.'

[Citation omitted.] In the report of the 'Panel on Poultry' issued on 21 November 1963, GATT Doc. L/2088, paragraph 10, the panel stated: '[T]he shares in the reference period of the various exporting countries in the Swiss market, which was free and competitive, afforded a fair guide as to the proportion of the increased German poultry consumption likely to be taken up by United States exports'. See also Panel report in 'Japan � Restrictions on Imports of Certain Agricultural Products, paragraph 5.1.3.7 [citation omitted]'."

38. It is to accomplish the chapeau's requirement that a "Member shall aim at a distribution of trade � approaching as closely as possible the shares which the various Members might be expected to obtain in the absence of restrictions", that Article XIII:2(d) requires, as one alternative, the allocation of shares on the basis of a previous representative period (adjusted for special factors if and to the extent appropriate).
39. If data from a period are out of date or imports distorted because the relevant market is restricted, then using that period as a representative period cannot achieve the aim of the chapeau. Thus, under GATT practice it is necessary that the "previous representative period" for purposes of Article XIII:2(d) be the most recent period not distorted by restrictions. As noted above, the panel on EEC - Restrictions on Imports of Apples from Chile, ¹⁸⁸ dealt with the question whether import restrictions reflected the proportion of imports to the European Communities "prevailing during a previous representative period" in the context of Article XI:2(c). That panel excluded the year 1976 from the most recent three-year period previous to 1979, the year when the EC restriction in dispute was in effect, and chose 1978, 1977 and 1975 instead. It held that 1976 could not be considered representative due to the existence of restrictions during that year.
40. The panel on Japan - Restrictions of Imports of Certain Agricultural Products ¹⁸⁹ addressed the question of the absence of a "previous representative period" in the context of Article XI:2(c). It noted that:

"� in the case before it the import restrictions maintained by Japan had been in place for decades and there was, therefore, no previous period free of restrictions in which the shares of imports and domestic supplies could reasonably be assumed to resemble those which would prevail today. � The Panel realized that a strict application of this burden of proof rule had the consequence that Article XI:2(c) could in practice not be invoked in cases in which restrictions had been maintained for such a long time that the proportion between imports and domestic supplies that would prevail in the absence of restrictions could no longer be determined on the basis of a previous representative period. � The Panel considered for these reasons that the burden of providing evidence that all requirements of Article XI:2(c)(i), including the proportionality requirement, had been met must remain fully with the contracting party invoking that provision."

41. We note that Article XI:2(c), which stipulates that quotas must be such as not to reduce the total of imports relative to domestic production which might reasonably be expected to rule between the two in the absence of restrictions, is an exception from the prohibition of quantitative restriction in Article XI:1. Article XIII regulates the non-discriminatory administration of quantitative restrictions, including, where applied, the allocation of shares among Members. The determination of a previous representative period under Article XIII raises similar problems as under Article XI:2. Thus we deem the above considerations pertinent to the case before us. The effect of a lack of a representative period under Article XIII is much less far-reaching than the lack of such a period under Article XI:2(c). In the Japan � Restrictions case, the lack of a suitable previous representative period precluded the use of the Article XI:2(c) exception. Under Article XIII, the lack of a suitable previous representative period would only preclude allocation of a tariff quota unilaterally. It would not preclude the use of a global tariff quota nor of country-specific allocations by agreement.

(b) The Representative Period

42. With regard to the selection of a "previous representative period" for applying the tariff-quota regime for imports of bananas to the European Communities, we recall that prior to 1993, EC member States applied different national import regimes. Some member States applied import restrictions or prohibitions, while imports to other member States were subject to a tariff-only regime or could enter duty-free. ¹⁹⁰ Thus, that period could not serve as a previous representative period (see paragraph 6.37).
43. With the introduction of the common market organization for bananas in mid-1993, we note traditional ACP supplier countries were guaranteed country-specific allocations at pre-1991 best-ever import levels, which were far beyond their actual trade performance in the recent past. As of 1995, the Banana Framework Agreement (BFA) allocated shares of the 2,200,000 tonne tariff quota established by Regulation 404/93 to the substantial suppliers Colombia and Costa Rica. Given the distortions in the EC market prior to the BFA, the shares assigned to Colombia and Costa Rica could not have been based on a previous representative period. Moreover, the BFA contained WTO-inconsistent rules concerning the export certificate requirements and re-allocations of unused portions of country-specific allocations exclusively among BFA signatories, which further aggravated such distortions. The shares of non-traditional ACP supplier countries were also distorted because of the country-specific allocations within the quantity of 90,000 tonnes that were reserved for non-traditional ACP suppliers.
44. It could be argued that within the "other" category of the 2,200,000 tonne tariff quota (autonomously enlarged by 353,000 tonnes as of 1995 for the EC-15), Ecuador, Panama and the non-substantial third-country suppliers without allocated shares were competing on a relatively undistorted basis during the period when the previous regime was in force (although less so after the BFA entered into force). However, given that, for purposes of applying the requirements of Article XIII, it does not matter whether imports from some supplier countries were relatively less distorted than others since distortions with respect to one (group of) supplier countries will have repercussions on the import performance of other substantial or non-substantial supplier countries within a single-product market.
45. Accordingly, in our view, the 1994-1996 period could not serve as a previous representative period because of the presence in the market of the foregoing distortions.
46. We also note that the world market excluding the European Communities is of limited value for purposes of calculating country shares based on a previous representative period because different banana-exporting countries have quite different market shares in different regions of the world. For example, Ecuador's world market share has increased from 26 to 36 per cent during the last decade and thus is significantly higher than its country allocation under the EC revised regime. ¹⁹¹ Panama had a world market share of approximately 2-3 per cent of the market outside the European Communities during the past decade which is much lower than its country allocation under the revised regime. The Philippines had a share of approximately 13-14 per cent of that market outside the European Communities during the past decade, but it does not export significant quantities to Europe. Thus, data on world-market shares of various supplier countries during any past period (regardless of whether such data includes or excludes exports to the European Communities) could hardly be relevant for purposes of calculating country shares based on imports to the European Communities reflecting a previous representative period. Because different banana-exporting countries have quite different market shares in different regions of the world, it would also be difficult, if not impossible, to use a regional or specific country market as a basis for allocating tariff quota shares.

(c) Special Factors

47. Ecuador suggests that the European Communities could comply with Article XIII by basing its system on the 1995-1997 period, with adjustments both for the need to cure the distortions that existed in the EC market and the changes in relative economic efficiency and competitiveness.
48. However, the European Communities did not use special factors to adjust the country-specific tariff quota share allocated to substantial suppliers under its new banana regime. While in theory special factors could be used to adjust shares based on a previous unrepresentative period so as to meet the requirements of the chapeau to Article XIII:2, at least in the present case it would be difficult to do so in practice. We recall that, according to the Notes Ad Article XIII:4 and Article XI:2 of GATT, "the term 'special factors' includes changes in the relative productive efficiency as between domestic and foreign producers, or as between different foreign producers, but not changes artificially brought about by means not permitted under the Agreement." We note that in the past, GATT dispute settlement panels have appraised the consideration of special factors, such as "an overall trend towards an increase in Chile's relative productive efficiency and export capacity � [as well as] the temporary reduction of export capacity caused by [an] earthquake". ¹⁹² In our view, however, it would be inconsistent with paragraphs 2(d) and 4 of Article XIII to take account of special factors with respect to only one Member (see paragraph 6.37).

(d) Ecuador's Country-Specific Tariff-Quota Share

49. The reliance by the European Communities on a previous unrepresentative period, and without adjustment for special factors, would suggest that Ecuador's country-specific tariff-quota share does not approach the share that it might be expected to obtain in the absence of restrictions, as required by the chapeau to Article XIII:2. This is confirmed by the significant growth over the past decade in Ecuador's share of the EC ¹⁹³ and world ¹⁹⁴ markets. This growth indicates that Ecuador's country-specific tariff-quota share is less than it should be under the rules of Article XIII:2.
50. While Members have a degree of discretion in choosing a previous representative period, it is clear in this case that the period 1994-1996 is not a "representative period". Accordingly, we find that the country-specific allocations assigned by the European Communities to Ecuador as well as to the other substantial suppliers are not consistent with the requirements of Article XIII:2.

To continue with Article I of GATT 1994

¹⁸⁶ The country-specific allocations for, e.g. Belize, Cameroon, Côte d'Ivoire and Jamaica seem to include allowances for investment made.

¹⁸⁷ Panel reports on Bananas III, paragraph 7.68.

¹⁸⁸ Panel report on EEC - Restrictions on Imports of Apples from Chile, adopted on 10 November 1980, BISD 27S/98, paragraph 4.8.

¹⁸⁹ Panel report on Japan - Restrictions on Imports of Certain Agricultural Products, adopted on 22 March 1988, BISD 35S/163, paragraph 5.1.3.7.

¹⁹⁰ For a description of the market, see panel report on EEC � Member States' Import Regime for Bananas, issued on 3 June 1993 (not adopted), GATT Doc. DS32/R, pages 3-7.

¹⁹¹ Ecuador's world market share outside the European Communities in different three-year periods were approximately as follows: 1988-1990: 25 per cent; 1990-1992: 28 per cent; 1993-1995: 30 per cent; 1994-1996: 32 per cent; 1995-1997: 36 per cent.

¹⁹² Panel report on EEC - Restrictions on Imports of Apples from Chile, adopted on 10 November 1980, BISD 27S/98, paragraph 4.17; panel report on United States - Imports of Sugar from Nicaragua, adopted on 13 March 1984, BISD 31S/67, paragraph 4.3; panel report on EEC - Restrictions on Imports of Apples from Chile, adopted on 22 June 1989, BISD 36S/93, paragraph 12.24.

¹⁹³ Annex II.

¹⁹⁴ Annex II.