A willing-to-hit, afraid-to-wound WTO study on the ATC?

Geneva, 12 Aug (Chakravarthi Raghavan) - The end of the WTO Agreement on Textiles and Clothing (ATC) and the phase-out of quotas and integration into the GATT rules of the textiles and clothing trade on 1 January 2005, will result in China and India gaining market shares to a significant extent in the EU, the USA and Canada, but the surge may be less than expected, according to a study published by the WTO secretariat.

Published with the usual disclaimers (about its not being interpreted as reflecting the WTO secretariat’s view), and as a discussion paper, the study is authored by Mrs Hildegunn Kyvik Nordas, a WTO staff economist, and is an estimation of outcomes based on an economic modelling, which has not been attached to the paper.

Overall, the study says that China’s market share in clothing in the US “after ATC” will be 50 percent, compared to 16% ‘before ATC’ (meaning 1997), and of India’s to 15 percent in the US compared to 4 percent at that time.

In the other major importing and restraining market, the EC, China’s clothing share is projected to be 29% after ATC, compared to 18% before the ATC, while India’s is projected to become nine percent in the EC after the ATC against 6% before the ATC.

Having tried to project dramatically, China garnering 50% of the market share because of the end of the ATC, and India as the second most important beneficiary (and at the cost of several of the African and some Asian developing countries) - and perhaps attempting to leave the impression that it may be justifiable to impose handicaps on these two Asian economies - the study attempts to soften this by saying: “countries close to the major markets are likely to be less affected by competition from India and China than has been anticipated in previous studies.”

Mexico, Caribbean, Eastern Europe and North Africa are therefore liable to remain important exporters to the US and EU respectively, and possibly maintain their market shares. This is even more likely given the preferential access they have to the markets through regional trade agreements.

The report then adds: “The countries that are most likely to lose market shares are those located far from the major markets and which have had either tariff free or quota free access to the US and EU markets or have non-binding quotas.  These countries will undoubtedly face adjustment challenges.”

With figures and tables, the study attempts to paint a dramatic picture of the consequences of the end of the over 40-year old discriminatory quota regime (originally began as the short, and then long-term cotton arrangement, and then became successive versions of the multifibre agreement, with each version resulting in more discriminatory restrictions covering more areas of textiles and clothing and against more countries).

Though there is some footnote reference, two of the important elements that over and above the quota regime have distorted the world trade in this sector over the last two or three decades, and more specially since the WTO and the ATC came into being, are not discussed nor is there any explanation as to why their effects on the trade have been ignored.

These two elements are: the rules of origin, used both in terms of the normal as well as preferential trade, and the further effects of tax/tariff regimes, in the treatment and customs valuation of imports of ‘outward processed’ clothing.  Clothing or apparel made up from textiles and yarn, imported from the restraining country (the US, EC or Canada) and exported after this ‘transformation’ to the markets from which the textiles and yarn have been imported, is subject to tax or tariffs only on the value added.

The tariffs, which even under a quota regime, is relatively higher than the average MFN tariffs in each of these markets, is calculated only on the ‘value-added’, which in clothing industry is mostly of low-skilled labour.

If a country ‘A’ has a textiles industry, and uses the output of that industry to convert into clothing and exports it to say the EU, it will be subject to a tariff on the C.I.F. value of the entire product. If however, country ‘A’ imports the textiles from the EU (as fabric or cut for apparel by designers), and then converts it into clothing, and exports the product back to the EU, the tariff applied is only on this value added in clothing.

The varying rules of origin in the importing countries, and the tax regimes associated with them, as also regional preferential arrangements have also had a role in the trade.

The study has figures on market shares as a result of the outcome of the trade ‘before’ and ‘after’ quota elimination (and in the figure on clothing for the US, it is put as ‘before ATC’ and ‘After ATC’.

Though the use of the term ‘After ATC’ in some of the figures used to dramatise the effect would give the impression that the simulated data applies to post-2004 end and/or 2005, in fact the simulations only show what would have been the respective shares in 1997 (the base year used for the model), if there had been no quotas.

The ATC (along with the WTO) came into being on 1 January 1995. But the quota regimes go back to 1962 (when the cotton textiles arrangements began), and then the 1973 Multifibre arrangements.

A section of the study explains how the ATC has been implemented, but is descriptive in how it has been implemented, in the letter (rather than the spirit?) and the backloading of quota phase-outs, adding, that in short the table (on integration in three stages) “leaves the impression that liberalization has been kept to a bare minimum.”

Why the assessment about liberalization has been left to be drawn on ‘impression’ in a simulated econometric modelling exercise is not clear.

The discussion paper, in looking at the impact of the phasing out of the ATC, and the outcomes of earlier studies used in static models, argues why static models do not provide a realistic picture of the outcome, and then explains the choice of the general equilibrium model for a simulation, using what is known as the GTAP model, arguing that it is “better at projecting relative performance of countries than absolute performance.”

The outcomes of general equilibrium models and simulation exercises are notoriously dependent on the various assumptions, and simulations based on change of one factor while useful for economists, can give a totally distorted picture. This is why the economics profession, even of mainstream economics, is divided on modelling exercises and recommending policies based on them.

However, the WTO study’s simulation, as well as the GTAP model (run by the Purdieu University in the US) and used in many econometric models are equally defective at least for the purposes of the study. The definition of textiles in that model, and used in the study is not the one commonly understood. The model lumps under textiles such items as T-shirts, singlets and other vests, jerseys, pullovers, cardigans, waistcoats and other such articles, panty hose, tights, stockings, socks and other hosiery.

Under the HS system of classification used at the WTO, and under its chapter 61 (articles of apparel and clothing accessories, knitted or crocheted - item nos 61.09 61.10 and 61.15) are clothing, and not textiles.

As a result the textile shares of some of the countries are probably overstated, and their clothing shares understated.

The GTAP model, used by the WTO, also includes leather apparel under textiles, whereas the WTO trade regime for leather products is quite different from textiles and clothing.

The projections in the discussion paper are based on modelling, and on page 25, the study says “The GTAP model with base year 1997 has been used, while the ATC was introduced in 1995 and the phase-out by 2005.”

The study then says, “as shown in section III, LITTLE had changed from 1995 to 1997. A simulation using 1997 as the base year should therefore not constitute a MAJOR problem for analysing the impact of the ATC.” (emphasis added).

Surely, some things did change between end 1994 (when the old MFA regime was there) and 1995 and 1995-1997.

To cite only one example, in implementing the Uruguay Round agreements, and the ATC integration process, the US changed its rules of origin, and the quota attributions of origin of clothing exported to it.

There was also (in the US context) the effect of NAFTA which came into being in 1994, and benefiting Mexico in textiles and clothing sector, and which began to bite in over 1995-97.

Could all these be ignored in assumptions, that taking 1997 as a base for simulation should not really affect or be a major problem?

And when all the figures present the picture as of ‘After’ ATC, is the outcome supposed to be that in 2005, overnight so to say (not very credible, even with the ‘flexible’ supply chains the study mentions) after end of quota regime, or after 2, 3 or 5 years.

And when the figure on page 30, on the market shares before and after clothing quota elimination in the US is presented (the model is based on US and Canada being aggregated as one region, though Canada has the same old rules of origin), surely the change in trade patterns as a result of the change in the US rules of origin do matter.

In its conclusion (para 3 of column1 of page 34), the study says that it has been able to take account of recent developments in the organization of the textiles and clothing sector, and the vertical specialization that implies inputs cross borders back and forth several times and trade is sensitive to tariff levels - and the prevailing tariff rates and preference margins.

Which of the market shares of countries in the figures is attributable to the end of the quota regime, and which to these developments? How have these figured in the modelling - inputs and outputs?

For example, while all Bangladesh exports to the EC can get GSP benefits, due to the peculiarities of the EC’s GSP rules of origin used for this, Bangladesh has been able to have a GSP use rate in the EC of only 39.6% in 2000, 46.4% in 2001, and 57.2% in 2002 (according to EU Commission data).

Or take the case of Mauritius, which gets AGOA benefits from the US, but can in fact use the privilege only if it imported textiles from the US (costlier than other sources) to re-export as clothing. Before AGOA, its exports grew at a faster rate than after AGOA.

However, Lesotho’s exports grew from a 12.06% growth rate over 1994-1999, it was able to increase its growth rate to 37.22% from 1999 to 2003.

The problem for Mauritius, and one that cannot be resolved by restraining the trade of China or India for example, is that with its per capita, under AGOA, it can benefit only by importing costlier textiles from the US, convert into clothing and export it to the US - to benefit by the rules of origin and the AGOA preferences. But this will make its apparel non-competitive. On the other hand, Lesotho as an LDC with a lower per capita has no such restriction.

Similarly, in the EC markets, Mauritius, though an ACP country, is not entitled to fully duty-free access which is available only to LDCs, and gets only 20% reduction in applicable duty.

However, Bangladesh exports to the EC as an LDC, and gets full benefits, with its exports increasing year after year, while that of Mauritius fluctuates.

The erosion of market share for a country like Mauritius (and countries similarly placed) is really due to the tariff preferences enjoyed by its competitors, and the rules of origin even where in theory preferences are available.

The study also explains some of the growth or benefits of some countries exporting into the US, as being due to their proximity to the US markets.

However, for example, Jordan is much closer to the EU than to the US. Despite this, its exports to the EU is just $12 million, while its apparel exports to the US have expanded from $49 million in 1999 to $583 million in 2003, and to $710 million in the year ending May 2004.

Surely, this is due to the duty free concession enjoyed by Jordan (under the US-Jordan free trade agreement), and the nearness to markets is not a factor, and ought to have been mentioned.

In Figure 10 on page 30, before the ATC(1997), Sri Lanka is shown as having no share in the US clothing market, but as having a 2% share after ATC.

On page 12, in table 32, the study presents what is said to be the ‘revealed comparative advantage’ of various countries, and this concept of ‘revealed comparative advantage’ is explained in a footnote No. 27 on the previous page 31.

This footnote explains that the “revealed comparative advantage is calculated as the share of textiles and clothing in total exports of each country relative to the share of textiles or clothing in world exports.”

“If the number is larger than unity, the country in question has a revealed comparative advantage,” the footnote adds.

On this basis, if one looks at the table, Bangladesh has a revealed comparative advantage of 18.63. China is then shown as having a revealed comparative advantage of 3.64 and for India it is 3.90.

Under the quota regime, China is depicted as having less than half of the clothing trade, under a binding quota; 7/21 share of India’s trade is under binding quota, while Bangladesh has no quotas but is under surveillance.

With such a high revealed comparative advantage, about five times that of China and India, could the end of the quotas be projected in the figures as responsible for cutting into the Bangladesh share? Or should it not be seen as perhaps due more to the rules of origin and the tariff preferences - and the problems viewed thus.

If the GTAP modelling cannot capture these, using it to study and present an analysis, without qualifications or explanatiions to point out its limitations is not sound trade economics. If the GTAP model can capture these elements, but has not been so used, the study becomes something different. –SUNS5636

[c] 2004, SUNS - All rights reserved. May not be reproduced, reprinted or posted to any system or service without specific permission from SUNS. This limitation includes incorporation into a database, distribution via Usenet News, bulletin board systems, mailing lists, print media or broadcast. For information about reproduction or multi-user subscriptions please contact: