5.90 The United States also replied to a question from India that the decline in production of 5,000 dozen units could be explained by a loss of export orders rather than an increase in import competition. The US rejected India's view that a given decline in production might have been the result of reduced export demand. CITA had found ample evidence of damage or the threat of damage occurring to US producers of woven wool shirts and blouses due to import competition and had received no information that there had been a decline in export orders. The United States also pointed out that, because of the relatively small number of woven wool shirts and blouses produced in the US, after rounding, the preliminary and final 1992 production data reflected in December 1994, April 1995, and currently were the same - 80,000 dozen.

Data on Exports

5.91 India argued that, to determine whether the share of imports of woven wool shirts and blouses into the United States' market rose or fell in 1994, it was necessary for the United States to collect data on exports that were comparable to those for imports and production. The United States Market Statement had not included such data.¹⁷ It was the responsibility of a Member that decided to impose a safeguard action to be in a position to provide all data relevant to an assessment of serious damage or actual threat thereof, and in particular exports. Otherwise, a safeguard action could not be taken consistently with Article 6 of the ATC. India had obtained figures on United States exports of woven wool shirts and blouses from official United States publications.¹⁸ According to these data, virtually all the United States production of woven wool shirts and blouses was exported, leaving imports to satisfy demand. This suggested that imports had satisfied a domestic market that had not been supplied by domestic producers and that changes in the level of imports could consequently not cause damage to the domestic industry.

5.92 In addition, India considered that, in order to determine whether, and to what extent Indian shirts and blouses were actually competing with US­made shirts and blouses in the United States market, the US would need to examine, inter alia, which portion of US production was sold domestically and which portion was sold abroad. The United States had refused to do so, claiming that its official export data were unreliable and that it could proceed on the basis of the "best data available". However, under the ATC, the United States must base its determination on a demonstration that it was the increase in imports and not other factors that had caused the serious damage and the United States must, therefore, collect the data necessary to make that demonstration. If the best information available did not include export statistics, while these statistics were necessary to make that demonstration, then the United States could not take the safeguard action.

5.93 In response to India's claims that the domestic industry could not be damaged by imports because domestic producers had chosen to export virtually their entire production, the United States explained that because of the known inaccuracy of the US export data, which had been pointed out at the time of the TMB proceeding in August 1995, official US export data could not be used to calculate the volume of the US market. The point made by India, that the entire production of the US woven wool shirt and blouse industry had been exported, was completely false and needed to be corrected. Official data on export quantities was highly suspect and could not be relied upon to assess conditions in the industry. Estimates by industry sources indicated that approximately 10 per cent of US woven wool shirt and blouse production had been exported.

5.94 The United States expanded upon the above points, explaining that it already knew from the two largest manufacturers in the industry that only 10 per cent of their production was exported. This information had been collected on a business confidential basis and no random sampling or scientific analysis was required or could be read to be required in the ATC. Nor was it necessary in this case where only 15 firms comprised the entire domestic industry. The data covering 60 per cent of the industry was excellent coverage and certainly CITA's reliance on this data was reasonable. In terms of exports, other sources were better than the official US data and this was also a problem with export data of other countries. India was wrong in stating that if it were true that there was a weak incentive to report export data accurately, this was true also of production and import data. In the United States reporting of production data from manufacturers was required by law and better reporting of import data was also required by law for duty collection and quota monitoring purposes in particular. This was not the case for exports.

5.95 The United States also explained that it had not provided India with the table referred to in paragraph 43 of India's first submission to the Panel; India had evidently developed the table on its own. The US reiterated that it had pointed out during consultations with India and during the August TMB proceedings that US export quantity data could not be used to calculate the volume of the US market because of the known inaccuracies of the export data. Even after all the shortcomings of the export quantity data were explained by the US in detail during the TMB review, India continued to use the inaccurate export data it obtained to incorrectly point out that the entire production of the US wool woven shirt and blouse industry was exported.

5.96 India reiterated that official US export data were available and was published not only by OTEXA, but also by the US Department of Commerce, Bureau of Census and the US Department of Agriculture. The detailed, and official export data of the United States allowed for a review of the quantity, value, and trends of exports of very specific and particular products including those that would be comparable to the import data contained in category 440. The facts, as presented in the official export data, indicated a clear decline in US export levels of products comparable to those in category 440. In particular, HS number 6205.30.15.00 identified exports of man­made fibre shirts containing at least 36 per cent by weight of wool and even if, as CITA contended, the data were not accurate, it at least indicated that a significant decline in the export of these products occurred between 1992 and 1994.

5.97 In response, the United States further advised that estimates obtained by CITA from the two largest individual domestic producers indicated that no more than 10 per cent of US woven wool shirt and blouse production was exported. If the market was adjusted for exports, assuming exports accounted for 10 per cent of the domestic production, the domestic market share in 1993 would decline from 53 to 51 per cent and for the first nine months of 1994 would fall from 40 to 37 per cent. As a result, the import market share in 1993 would increase from 47 to 49 per cent and for January-September 1994 would increase from 60 to 63 per cent. More generally, in characterizing the US data as unreliable, India was apparently contending that an importing Member could not resort to its ATC Article 6 rights to take a safeguard action without first obtaining all of the data necessary to respond to any conceivable challenge the exporting Member might make, and that all of these data must be publicly available. Acceptance of this argument would require that the data presented by importing Members in justification be limited only to information obtainable from public sources, however limited or inapplicable that information might be. In fact, there was no such limit in the ATC.

5.98 In summing up its argumentation, India claimed that the responsibility for compiling, examining and supplying to the exporting country the relevant data in respect of factors referred to in Article 6.3 of the ATC was entirely that of the importing Member. In the present case, the US had not supplied any information to India either in the consultation request or during the consultations, relating to one very important element to determine the state of the US industry vis­à­vis the exports effected by the US industry in category 440. India had collected US export data from the figures published by the US Department of Commerce. The US termed its own published data as "inaccurate" and "unreliable" but were not in a position to furnish any more reliable and accurate export data. If the published US data could not be used to assess the volume of US exports then there was no other way of correctly doing so. Different figures on production and exports as published by the US have been tabled by India and these figures have shown that a quantity equal to the entire production of the US in category 440 was exported. The US presentation claimed that the published official data of the United States on export quantities was highly suspect and suggested that "estimate by industry sources indicate that approximately 10 per cent of wool shirts and blouses production is exported". India submitted that the estimate of approximate quantities by the industry sources could not be held more reliable than the official data published by the US.

5.99 In the view of India, the US submission also failed to explain whether the exports of man­made fibre/wool blended shirts containing between 50 per cent and 64 per cent of man­made fibre had been taken into account while estimating wool shirt and blouse exports, while for import purposes, these were considered under category 440. It was India's understanding that in the absence of any specified procedure for culling out the export data, these were classified as man­made fibre shirts for export purposes. US data on the export of these blended fibre shirts had been submitted to the Panel in India's response to questions on 20 September 1996 and these data had shown that the entities exporting these products had experienced a significant decline in 1994 whether reported in dollars, dozen, or raw fibre equivalents. It was, therefore, more than reasonable to assume that this decline in exports would have more of an impact on the industry data supplied by the United States than any increase in imports.

5.100 The United States also summed up its position which had consistently been that US export quantity data was unreliable and could not be used in assessing conditions in the US industry. India had persisted in using this flawed evidence not only to support its untrue assertion that most of US production in category 440 was exported but also to denigrate the US production data and market share calculations. The deficiency of the export data stemmed from the low incentive of exporters to properly report the data and the absence of procedures to verify its accuracy. As pointed out in an attachment to the first US submission to the Panel, the Trade Data Division of OTEXA and the Bureau of the Census conducted an investigation of US exports of woven wool shirts and blouses and found that in 53 of the 201 exportations, the quantity reported was either zero or unrealistic; the Census Bureau talked with two US exporters who said they exported clothing but had no idea of its fibre content. The 6-digit Schedule B number was reported incorrectly in 4 of the 6 records examined and the correct Schedule B number could not be determined. More recently, in response to questions raised by India regarding US exports under Schedule B number 6205.30.1500, men's and boys' shirts of man­made fibre, containing 36 per cent or more by weight of wool, the Trade Data Division had conducted a shipment­by­shipment investigation of this export data. This investigation covered 1994 shipments of 7,554 dozen shirts which were made in 32 separate shipments. Most of the shipments were small and from different companies to different countries. However, four shipments were made by the same company to Honduras and represented 51 per cent of the total exports in this particular Schedule B number, i.e. 3,840 dozen. The Trade Data Division requested the Foreign Trade Division of the Bureau of the Census to review the data reported in these shipments. They found that all the shirts in these export shipments were in fact cotton woven shirts and were incorrectly classified. The 3,840 dozen shirts should have been classified under Schedule B number 6205.20.3000, men's and boys' woven shirts of cotton, not 6205.30.1500, for wool.

5.101 In the United States' view, the results of this investigation supported OTEXA's previous investigations and determinations that US export quantity data were not reliable. India's assertion that the estimate of approximate quantities of exports obtained by "industry sources cannot be held more reliable than the official data published by the United States Government" was wrong. Investigations conducted by OTEXA and the Bureau of the Census clearly indicated that US export quantity data were unreliable and inaccurate, making this information unsuitable for analytical purposes. CITA had, as mentioned above, obtained estimates from the two largest individual domestic producers of woven wool shirts and blouses, representing at least 60 per cent of domestic production, and they had indicated that no more than 10 per cent of US woven wool shirt and blouse production was exported. There was no basis to contend that information specifically requested from and supplied by companies about an important component of their sales would not be more reliable than unverified data that had been proven incorrect.

5.102 India argued that the calculation of export levels should have been made on the basis of reliable data when the determination of serious damage was made, and not subsequently in response to a query by India in the context of a panel proceeding. The recalculations of the United States only served to highlight the point made by India that export data were essential for the calculation of market share, and that data and other information used in a determination of serious damage must be verifiable to constitute the basis of the demonstration required under Article 6 of the ATC.

5.103 India also noted the United States had claimed that it was consistent with the requirements of Article 6.2 and 6.3 of the ATC to collect data on total production by directly contacting the producers benefitting from the safeguard action and at the same time it claimed that the data on the exports of domestic production was not available because the official export statistics were not reliable. It was questioned why the United States considered it consistent with the ATC to collect the information favourable to domestic producers (total production) informally through direct contacts, but it was only after its determination of serious damage that the United States informally contacted two of the fifteen producers to obtain information on the share of its production exported. Why was this not done before making the determination?

Data on Employment¹⁹, Man-hours and Wages

5.104 India pointed out that, with respect to employment, man-hours and total annual wages, the information provided in the Market Statement referred to the "748 establishments in the US that manufacture woven shirts and blouses including shirts and blouses made from wool". The Statement made reference to the fact that "employment in the industry producing woven shirts and blouses including shirts and blouses made from wool had declined to 31,929 production workers in 1994, six per cent below the 1993 level and a loss of 2,125 jobs". If the loss of 2,125 jobs was placed in relation to a decline in United States production of woven shirts and blouses from wool of 5,000 dozen between January-September 1993 and January-September 1994, it implied that a decline in production of 3 dozen woven shirts and blouses on an annual basis led to the loss of one job, clearly an absurd inference. The Statement went on to claim that "the average annual man-hours worked dropped" and total annual production worker wages fell", even though both claims referred to the industry producing all woven shirts and blouses, not the portion producing woven wool shirts and blouses. The fact that data on the industry producing all woven shirts and blouses was entirely irrelevant to the sub-sector making woven wool shirts and blouses was confirmed by the information submitted by the United States to the TMB in August 1995 under "Other Relevant Information". In the August 1995 submission, it was made clear that 200 workers were employed in the production of woven wool shirts and blouses in 1994, as compared to 215 workers in 1993, a total loss of 15 positions. India considered that data on employment, wages and man­hours at a more disaggregated level for the specific industry producing woven wool shirts and blouses should have formed part of the Market Statement provided to India as the basis for the consultations in April 1995, and were requested by India at the time. Employment figures submitted by the United States in April 1995 and August 1995 were:

5.105 India also noted that the data on employment in the wool shirt and blouse sector was specifically requested during the consultations and the Indian delegation was informed that such data did not exist. However, the data on employment in the wool shirt and blouse industry was included in the so­called "other relevant information" provided to the TMB on 28 August 1995.

5.106 India further pointed out that if the figures on employment provided by the United States in August 1995 were placed in relation to those provided by the United States in its April 1995 Market Statement, they indicated that the wool sub-sector accounted for 0.6 per cent of employment in the woven shirt and blouse industry. Since the sub-sector was an extremely small, if not negligible, portion of employment in the domestic woven shirts and blouses industry, the figures provided by the United States on employment, man-hours and wages in its Market Statement were totally irrelevant.

5.107 With respect to the above point, the United States claimed that it had indicated to India during consultations that employment data relating specifically to category 440 were not available, meaning only that such data could not be obtained directly from published sources nor was it regularly compiled for CITA. Data on employment and wages were published only at a higher level of aggregation than the woven wool shirt and blouse industry and at the time of the request the data given in the Market Statement was the most detailed that CITA was able to provide. It was not true, as India was implying, that the United States deliberately withheld such data from the Indian delegation during consultations. In actuality, when it became apparent that the justification for the request was being questioned by India because of the lack of this data and after indications from the TMB that such data would be a necessary element of their consideration of the case, CITA pursued ways of developing the requested information. Only after developing a methodology to further disaggregate the available data was OTEXA later able to provide, at the insistence of India during consultations and in accordance with the wishes of the TMB, more specific estimates based on additional information obtained from official and industry sources, which confirmed the downward trend of the broader category data reflected in April 1995.

5.108 In response to the points raised by India, the United States commented that it was correct that the employment-related information from the Market Statement was applicable to the industry producing woven shirts and blouses. CITA believed the more aggregated data to be generally indicative of the trend in the woven wool shirt and blouse industry at that time and received information from industry sources confirming this fact. OTEXA was later able to provide more specific estimates based on additional information obtained from official and industry sources.

5.109 India commented that the US had not explained why the more aggregated data of the total woven shirt and blouse industry was not indicative of the trend in terms of production, prices, profits, exports, imports, or any of the other relevant economic variables that should be reviewed prior to making a determination of serious damage or actual threat thereof. The increased production in the total woven shirt and blouse industry appeared to be ignored at the aggregate level because it contradicted the conclusion made by the United States concerning the trend in production data.

5.110 The United States referred to the above claim of India that the figures supplied by the United States on employment, man-hours, and wages were irrelevant because they covered the entire woven shirt and blouse industry and not just the woven wool shirt and blouse industry. The US pointed out that employment data presented in the Market Statement encompassed the entire US woven shirt and blouse industry and were derived from official Bureau of Labour Statistics (BLS) data covering even higher clothing production aggregates. This was the best information available at the time of the request for consultations. As a result of questions during consultations mandated by Article 6 of the ATC and as indicated by the TMB, the United States had provided the TMB with a breakdown of employment for category 440, woven wool shirts and blouses. CITA believed the more aggregated data to be generally indicative of the trend in the woven wool shirt and blouse industry at that time and received information from industry sources confirming this fact. CITA did not look at the trend in production for the woven shirt and blouse industry since CITA already had production data relevant to the wool sector of this industry.

5.111 The United States pointed to India's claim that because later data had shown that the number of jobs lost in the woven wool shirt and blouse industry was estimated at only fifteen jobs, there was no basis for the US determination of serious damage or actual threat thereof. In this regard the US recalled that the domestic industry in category 440 was very small, representing only 15 firms. Even though the loss of 15 jobs may, at first glance, appear small in absolute terms, it represented almost a 7 per cent decline in the number of production workers in one year. It would be difficult to argue that this was not a significant relative loss of employment. Furthermore, the United States found no indication in the language of Article 6.2 and 6.3 of the ATC that the term "domestic industry" was reserved for larger groupings of companies with greater numbers of workers. Indeed, the language of Article 6.2 of the ATC referred to "safeguard action" and the "domestic industry producing like and/or directly competitive products." This language placed no legal barriers on the maintenance of a safeguard action where the product may be narrowly defined or the industry small.

5.112 In response to a question from India asking if the number of production workers certified as eligible to apply for Worker Adjustment Assistance (220 workers) was more than nine times the decline of production workers (24 workers) during the April 1993-April 1995 period, the United States explained that the 220 workers employed in facilities producing woven wool shirts and blouses, that were certified as eligible for Workers Adjustment Assistance during the two and a half year period, January 1993-July 1995, included production workers as well as those workers employed in administrative, sales, and distribution positions associated with such production. Not all workers certified as eligible for Workers Adjustment Assistance had permanently lost their jobs; in many cases, workers were partially separated or temporarily laid off. (See also paragraphs 5.157 to 5.159.)

Information on Prices

5.113 India questioned if the information on domestic prices in the Market Statement could be considered to be representative of the situation of the particular segment of the industry producing woven wool shirts and blouses. According to the Market Statement, the industry statements were "based on information supplied by individual US firms domestically producing shirts and blouses", and "in general ... applies to companies producing men's and women's woven wool shirts and blouses". In other words, the information had been obtained from enterprises that manufactured woven wool shirts and blouses as part of their production of woven shirts and blouses. It was also questioned if it was appropriate to use informal surveys of enterprises as the basis for taking an action against the imports of a trading partner. During the bilateral consultations held in April and June 1995, India's delegation had sought clarifications from the United States' delegation regarding the underlying methodology that had been used. The United States delegation' had confirmed that there was no procedure for regular or periodic compilation of price data. Data relating to price and the disaggregation of employment data for specified product segments such as woven wool shirts and blouses etc. were based on informal surveys of a limited number of firms producing these items. There was no scientific random method or a stipulated sample size for such surveys. It was also noted that the firms responding to such surveys were always aware that the purpose of the survey was to initiate a safeguard action to protect that segment of the industry.

5.114 In the view of India, the informal methods used to survey enterprises might explain the wide variations of the results of such surveys reflected in the different industry statements furnished by United States. For example, in the December 1994 request for consultations, the average producers' price was reported as $215-225, while in the Market Statement, the average producers' price was reported as $525-550. Since it was unlikely that producer prices would double in such a short period of time, this discrepancy between the two Statements by the United States cast doubt on the consistency of information collected by informal surveys.

5.115 The United States explained that the difference in the two prices was not caused by an increase in domestic prices; rather, the two prices represented the average prices of two different groups of products. Prices as reported in the December 1994 market statement for category 440 under the MFA reflected the average domestic producer prices for wool shirts comparable to wool shirts imported from India which were concentrated in one of the 24, 10-digit product classifications in the Harmonized Tariff Schedule of the United States Annotated (HTSUSA) that made up category 440. This had been done because MFA determinations were based on sharp and substantial increases in products by country. The US had compared the average, landed duty-paid value of wool shirts imported from India classified under HTSUSA 6205.10.2010 - men's wool shirts, other than hand loomed and folklore shirts - with the average price of domestically produced men's woven wool shirts. The average US producers' price in the Market Statement issued in April 1995 under the ATC represented the average domestic price for all woven wool shirts and blouses produced in the US which competed with all woven wool shirts and blouses imported from every country in category 440. Under the ATC, the initial determination was on total imports in the category. Therefore, the $525/$550 per dozen average import price in the Market Statement was examined based on the United States' reading of Article 6.2 of the ATC requirement of an examination of "total imports". By contrast, the $215/$225 average import price in December 1994 was based on a particular product from a particular country (i.e. India) which was the analytical approach required by the MFA.

5.116 The United States, in response to a question by India whether the substantial price differences could be explained by quality differences (low priced imports and high priced domestic production), responded that the average landed duty-paid import value for total US imports of category 440, woven wool shirts and blouses, was $US 187.23 per dozen while such imports from India were valued at $US 133.85 per dozen or 75 per cent below the average US producers' price for domestically produced woven wool shirts and blouses, and 29 per cent below the category 440 average landed duty-paid value for total US imports of woven wool shirts and blouses. The price difference between domestically produced woven wool shirts and blouses and imports (including those from India) was primarily the result of differences in labour costs that varied among all countries producing woven wool shirts and blouses. Quality differences reflected in prices of woven wool shirts and blouses included differences in hand tailoring, the quality of wool fabric, fibre content, fibre blending, detail included, etc. which varied among all countries that produced woven wool shirts and blouses. The domestic price for woven wool shirts and blouses reflected the average price of all domestically produced woven wool shirts and blouses and was compared with the average landed, duty-paid import values at the category level (all products imported in the category) from each country supplying the US market and the average import value for all supplying countries. The United States did not accept India's assumptions that in a single market prices of competing products would "normally tend to converge" or that products of different quality and which were sold at varying retail prices could not "compete".

5.117 In response to the above point, India noted that the US had established a number of "quality differences" for these woven wool shirts and blouses, but offered no data on the various quantities that were produced among these various quality differences. It would have been interesting to see the trend in production of those shirts which in December 1994 were at $225 per dozen for comparable shirts being imported from India, whereas the average US price for all woven wool shirts and blouses was $550 per dozen. This would have indicated that not only was there a wide quality difference among the shirts produced in the United States, but also that those shirts which were directly comparable and competitive with the shirts from India may have increased or producers may have shifted to the higher value shirts. There must have been some discrimination in the presentation of price and production data that would indicate that US data were comparable to those products in category 440 which were claimed to be seriously damaging or actually threatening serious damage to US producers of "like and/or directly competitive products".

5.118 The United States noted in this regard that quality differences, as reflected in prices of woven wool shirts and blouses varied among all countries that produced woven wool shirts and blouses. The domestic price for woven wool shirts and blouses reflected the average price of all domestically produced woven wool shirts and blouses and was compared with the average landed, duty­paid import values at the category level from each country supplying the US market and the average import value for all supplying countries. Domestic producers of woven wool shirts and blouses in the relatively narrow category 440 competed with imports from India and from all the other suppliers.

Information on Investment and Capacity

5.119 India noted that the Market Statement had included information on investment and utilization of capacity based on industry statements. As follows:

5.120 This information was, in the opinion of India, anecdotal and unverifiable. It was also unclear whether the information referred to the particular segment of the woven shirt and blouse industry producing garments made from wool. For example, the fact that "several companies reported a decline in capacity" did not appear to be significant in the context of an industry reported by the United States in its Market Statement as comprising 748 establishments. It was also argued that one company reported dropping of contracts or reduced capacity utilisation which was not an appropriate indication of the capacity utilisation for the entire industry. If the production capacities of several companies that had actually declined were related to the wool shirt and blouse industry, the decline in domestic production during 1994 should have been much more than an estimated 8 per cent. Other information provided in the Market Statement, (and reproduced below) was equally anecdotal and unverifiable. The information on "profits" was in fact on "profit margins", leaving it unclear whether total profits had declined or increased. India argued that the United States had provided no proof for the assertions in the industry statement regarding the role of "lower-priced" imports in industry developments.

5.121 In response to India's assertion that one company reporting dropping of contracts or capacity utilization was not an appropriate indication of the capacity utilization for the entire industry, the United States argued that, given the small size of the woven wool shirt and blouse industry, the decline in capacity utilization from this one company alone was highly indicative of what was going on in the entire woven wool shirt and blouse industry. India had also alleged that the decline in domestic production during 1994 should have been more than an estimated 8 per cent; however, the loss in capacity utilization did not necessarily correlate with a commensurate drop in production during the same time-period. Rather, the loss in capacity utilization was an indication of deteriorating conditions in this industry that would lead to more severe production declines in the future.

F. Causal Link Between Increased Imports and the Domestic Industry Situation

5.122 According to India, the Market Statement submitted by the United States in April 1995 stated that "the sharp and substantial increase in imports of woven wool shirts and blouses, category 440, is causing serious damage to the US industry producing woven wool shirts and blouses". Since the figures on market share, employment and wages were, as argued by India, irrelevant, and the figures on domestic prices and information on other relevant economic variables were based on questionable survey methods and were unverifiable, the only real evidence provided by the United States in support of its assertion of serious damage was the fact that imports of category 440 had increased in 1994 by 69,296 dozen to nearly double the previous year's level, while domestic production had dropped marginally by 5,000 dozen during January-September. While it had been claimed by the US that production had declined due to imports, no analysis was provided to link the two. Nor was the decline in production proportionate to the increase in imports. In the industry statement, there were claims of loss of employment, closure of plants, loss of profits etc. arising from imports; however, no attempt had been made to link these developments to imports. The Market Statement submitted by the United States in April 1995 never went beyond assertions.

5.123 The United States argued that the causation requirement in Article 6.2 of the ATC, linking serious damage, or the threat thereof to total imports, had been met in this case. As evidenced in the information provided in the Market Statement and later to the TMB: (i) imports had not only increased, but surged; there were negative industry indicators occurring contemporaneously with those surging imports; (ii) about 7 per cent of the workers in the woven wool shirt industry had lost their jobs from 1993 to 1994 (from 1994 to 1995 there was a loss of 5.9 per cent); later data supported this trend that the adverse impact of imports on employment was evidenced by the US trade adjustment assistance certifications (by US law a connection has to be made to imports to be eligible for certification); and (iii) US market share had declined as imports increased and production declined at the same time that imports increased.

5.124 In the view of the United States, CITA had demonstrated in the Market Statement and at the TMB proceeding the causation required under the ATC. Although India has asserted on this issue that "positive evidence" was required, the United States found no evidentiary standard in the ATC and could only conclude that India was adding to the text of the ATC provisions that were not negotiated and were not intended as an interpretation of the Agreement by the US.

5.125 The United States considered that India was seeking to modify the ATC by creating a proportionality requirement to establish a causal link. India had claimed that the United States must demonstrate that the decline in production evident before CITA was "proportionate to the increase in imports". The United States found no such test in Article 6.2 or 6.3 of the ATC. Nor was there a factual or economic justification that would require a finding that serious damage to the domestic woven wool shirt and blouse industry by imports would be reflected by exactly proportional changes in production and imports. The United States' imports of woven wool shirts and blouses from a number of countries were limited by quotas. There was also a sharp seasonal variation in these imports as well as differences in the timing of production and import activity.

5.126 India stated that it had never proposed a proportionality requirement, but had remarked that US production had never varied previously with imports and that this lack of correlation suggested that factors other than imports must have influenced the level of domestic production and, in particular, developments on the export market. India agreed that factors such as sharp seasonal variations in imports and differences in the timing of production and import activity made it impossible to conclude from the mere co­existence of rising imports and declining production that the two were causally linked.

5.127 India further argued that a demonstration that there had been a rise in imports and a decline in production was not a demonstration that there was a causal relationship between the two; logically, additional facts and data were necessary. Article 6.2 of the ATC explicitly stipulated that a demonstration that there had been an increase in imports and serious damage or threat thereof was not sufficient but must be supplemented by an additional demonstration that the increase in imports and not other factors were causing the serious damage or actual threat thereof. This demonstration of causality had not been attempted by the United States.

5.128 Concerning the lack of a causal link, India referred to the data on the dollar value of exports submitted by the United States to the TMB in August 1995 which indicated that the value of exports of woven wool shirts and blouses from the United States had increased by 41 per cent in 1993 and by nearly 30 per cent in 1994. Since a major portion of domestic production of woven wool shirts and blouses was exported, the domestic industry was, in fact, experiencing a significant improvement in the period prior to the imposition of the safeguard action in July 1995. Also, in its submission to the TMB in August 1995, India had pointed out that imports from India of category 440 were steadily dropping in 1995. This statement was confirmed by figures on imports submitted by the United States. During the first six months of 1995, imports from India had amounted to 2,887 dozen, 67 per cent below the earlier year's figure. The condition of increased imports was therefore not met in July 1995 when the United States unilaterally imposed restraints on imports of category 440 from India. During the period 18 April 1996 to 2 August 1996, i.e. the first three months of the second year of the continuation of the restraints, the actual imports from India had been less than one per cent of the restraint level imposed by the United States. Thus the subsequent data and import statistics proved beyond any doubt that the attribution of actual threat of serious damage to the domestic industry to imports from India had been grossly misplaced and the finding of the TMB on this point was, therefore, wrong.

5.129 India further argued that the absence of a causal link between increased imports and declining production of woven wool shirts and blouses was demonstrated by figures over a longer time period. From 1985 to 1992, imports of woven wool shirts and blouses had fallen consistently and substantially, from 262,000 dozen in 1985 to 44,000 dozen in 1992. During the same period, production had also declined substantially, from 445,000 dozen to 80,000 dozen. Thus, declining production was accompanied by declining imports for the period 1985-92. In 1993, the United States market for woven wool shirts and blouses had begun to recover, with both production and imports rising. In 1994, imports nearly doubled, while production declined by 7.5 per cent. Since a major portion of United States production was exported, the increase in imports was obviously related to expanding domestic demand for woven wool shirts and blouses. The United States industry producing woven shirts and blouses had probably not anticipated this development in the wool segment since the market had been declining for a number of years. One explanation for a lack of correlation between imports from India and US domestic production was that Indian and US products were not actually competing with one another in the US market because they fell into different price and quality categories. Another possible explanation was that, while India supplied the US market, the US producers supplied both the domestic and the export markets. India's exports, therefore, varied solely with the demand in the US market; US production varied also with the demand in other countries.

5.130 India noted that, in respect of the import data going back to 1983, these data related to the imports and production of products defined in category 440, were derived by the United States Department of Commerce, OTEXA, and were contained in their periodic publications. The import data for 1983 and 1985 were the total reported imports of woven wool shirts and blouses that were in chief value wool while the reported imports from 1989 through 1994 were for woven shirts and blouses in chief weight wool and in chief weight man-made fibre if they contained 36 per cent or more by weight of wool. There was no publicly available data for production and it was assumed that this data relating to category 440 by OTEXA was derived by that agency for use by CITA in assessing the US domestic market for these products.

5.131 The United States considered that the reference by India to production and import data going back as far as 1983 was an effort to deflect attention away from the surge in imports from India that had occurred in the time leading up to the issuance of the request for consultations. India's proposed time series dating back to 1983 was technically flawed as the production data cited covered a time period that included two census survey benchmark years, 1987 and 1992. Data prior to those years were not comparable to the subsequent years' data due to differences in the composition of the survey sample. Import data were likewise not comparable over the period of years given, because, beginning in 1989, the United States had shifted to the Harmonized Tariff System classification. This shift involved a change in the wool shirt definition from a "chief value wool" basis to a "chief weight wool" basis, that caused the data prior to 1989 to be not comparable with subsequent years' data. On India's assertion that a major portion of US production was exported and its subsequent assumption that as a result the domestic industry was in fact experiencing improvement, the United States noted that the assertion and the assumption were false because export data was extremely unreliable. The United States had already illustrated that point by confirming that only 10 per cent of the production was exported by the industry representing 60 per cent of US production and that there were misclassifications of cotton exports under the wool heading.

5.132 India considered that the US' view in the first sentence of the preceding paragraph was a misrepresentation of the arguments made by India, which referred the Panel to the absence of a consistent relationship between changes in imports and changes in domestic production. This argument would remain valid irrespective of the shift in the United States data collection methods. For instance, between 1985 and 1989, a period in which data on imports and production were presumably collected on a consistent basis, both imports and production declined substantially. Starting from 1990, there was again no consistent pattern between changes in imports and domestic production, undoubtedly reflecting developments in the domestic and export markets. It was for this reason that India had asked for information on exports.

5.133 The United States argued that India's assertion that between 1985 and 1989, data on production had been collected on a consistent basis was not true. As stated in the US' first submission, India's data covered a time period that included one census survey benchmark year, 1987. Data prior to that year was not comparable to the subsequent years' data due to differences in the composition of the survey sample.

5.134 In response to these arguments, India submitted that the reliability or comparability of production data could not differ significantly from one census survey to another if all the census surveys were objectively and scientifically done. In assessing the role of imports in influencing or not influencing the production trends, it was not only justified but essential to look at the relation between imports and production during as long a period as possible. These data established the fact, over a significant period of time, that changes in import levels were not correlated proportionately or otherwise to the changes in domestic production. In terms of the US shift to the Harmonized Tariff System, the US submission was factually incorrect. The wool shirt definition prior to the Harmonized Tariff System had been based on a chief value determination. Under the Harmonized Tariff System, the classification was based on a chief weight determination, but the US had developed statistical breakdowns for both exports and imports to identify shirts and blouses in chief weight man­made fibre but containing 36 per cent or more by weight wool. The statistical breakdowns were developed and implemented in order to identify those woven shirts and blouses in chief weight man­made fibre, but, on the basis of estimates by the United States, similar to those woven shirts and blouses that were in chief value wool. Thus, these chief weight man­made fibre shirts and blouses were included as part of category 440 even though they were in fact man­made fibre shirts and blouses.

5.135 In response, the United States argued that India's assertion was incorrect that the reliability and comparability of production data could not differ significantly from one census survey to another. First, there was no issue of reliability but only of comparability. Comparability was lost because, in the process of revising the five­year Census of Manufacturers, new firms were identified and a new sample and sample size was established which included a different group of firms than the previous survey. As a result, the five­year Census of Manufacturers established a new benchmark, and data generated by this new survey were not directly comparable with previous years' data which were generated from reports of the old sample of firms. The Bureau of the Census production data in its quarterly Current Industrial Report (CIR) were based on data collected from firms identified in the five­year Census of Manufactures. Starting with production data collected for 1992 in the CIR, the number of firms originally identified in the 1992 Census of Manufactures was revised every year with the Annual Survey of Manufactures, which was taken during the intermediate census years. The annual revisions to the sample size were reflected in the CIR production data. This data collection process was not in effect prior to 1992.

TO CONTINUE WITH USA - MEASURE AFFECTING IMPORTS OF WOVEN WOOL SHIRTS AND BLOUSES FROM INDIA

¹⁷ In its August 1995 submission to the TMB, the United States provided data on the dollar value of exported wool woven shirts and blouses. These data cannot be compared to the data provided on imports and production. The reason is that the export data supplied by the United States is in value (dollar) terms while the data on imports and production is in quantity terms (dozens). The United States explained in a footnote that export quantity data are questionable due to reporting inconsistencies.

¹⁸ 1993: Production 82,000 and Exports 85,000. 1994: Production 76,000 and Exports 76,000.

¹⁹ See also paragraphs 5.154 to 5.156.