Tender 06-H13. Study on the Application of Value Criteria for Textile Products in Preferential Rules of Origin FINAL REPORT

Transcription

1 Tender 06-H13 Study on the Application of Value Criteria for Textile Products in Preferential Rules of Origin FINAL REPORT Prof. Dr. Michiel Scheffer Saxion Universities Enschede With the contribution of Mr. Klaas Winters Prof. Dr. Ali Bayar For Limited Circulation Enschede, December 20th 2006 Opinions expressed do only engage the consultant and do not in any way reflect the views or opinions of the European Commission.

2 TABLE OF CONTENTS INTRODUCTION... 4 METHODOLOGY PREFERENTIAL TRADE AND RULES OF ORIGIN Preferential Trade and Rules of Origin Architecture of Preferential Rules of Origin Sufficient Working or Processing Cumulation Tolerance Rules Territorial Continuity Absorption Principle Discussions on Reforms of Rules AGOA Canada s Market Access Initiative for LDC s EU Variations on Double Transformation The Academic Discussion on EU Rules of Origin Conclusions TRADE DYNAMICS Trade Patterns Take Off Ceiling Stupor Maturity Do Origin Rules Make a Difference? The EU as a Donor Country Conclusion SUPPLY & VALUE CHAIN The Textile Supply Chain Calculating Value Criteria Results of Calculation Model % Criterion % Criterion Modulation by Product Group Distribution of Costs Distribution Analysis The Sri Lanka Costings Modulation by Country Analysis of Trade Flows in Relation to Valuation Product Modulation Country Modulation Conclusion SCENARIO AND MODELISATION ASSESSMENT Scenarios Scenario 1 Reviewed Value Criteria Scenario 2 Simple Transformation Value Criteria Scenario 3 Modernized Double/Single Transformation Limitation of the Model Results of Simulations Evaluation COUNTRY CASES

3 5.1 Sri Lanka Competitive Context Strategy of the Industry Trade Patterns The GSP System for Sri Lanka Evaluation Lesotho Garment Industrialisation in Lesotho Vertical Integration Availability of Textiles in Africa Evaluation Tunisia Structure and Strategy of the Industry Trade Patterns The Preferential System for Tunisia Evaluation The European Union Major Trends Bottlenecks in Current Rules Evaluation IMPLEMENTATION AND QUALITATIVE IMPACTS Which Value Criterion Strategy Impact of Value Criteria Technological Choices Specialisation Choice in Materials Upgrading Commodization and Efficiency Tactical Consideration Negotiations Fluctuations Fluctuation in Prices for Commercial/Industrial Reasons Implementation Issues The Exporter The Importer Conclusion EVALUATION ANNEX 1 CALCULATION CHART ANNEX 2 COMPANIES INTERVIEWED OR COMPANY DATA USED ANNEX 3 A DESCRIPTION OF THE GREENMOD MODEL

4 INTRODUCTION A micro fibre bra is being made in Sri Lanka. It is a garment containing 35 parts, some are textiles like the straps, the lace, and the elastic knitted fabric in the front and back, some are metal or plastics, like the clips and braces. Making a bra is labour intensive. It requires quality control at the beginning and end, precise logistics inbound and outbound. The value added in a bra is 50%. All but 3 textile components come from Sri Lanka. These three components of Thai origin represent 13% of the value of the garment. For that reason the product is not eligible for preferential import into the EU. Interview with manufacturer in Sri Lanka. A grey fabric, made of cotton/polyester is imported from Pakistan under the GSP. It is dyed and finished in the EU in order to make in the corporate identity of the client. The fabric receives a flame retardant treatment as well as a water repellent finish. The grey fabric bought for 1,20/meter is sold for 2,75/meter to a client sending the fabric for making up in the Former Yugoslav Republic of Macedonia. Although the fabric had preferential status when imported into the EU and although the fabric has undergone substantial added value in the EU, it cannot be considered as having preferential status after making up, while trade is also preferential between FYROM and the EU". Interview with Belgian manufacturer. These examples, collected out of many in this survey, illustrate the effect of the current rules of origin determining import of textiles from developing into the European Union. They demonstrate the complexity emerging from a patchwork of preferential agreements and the effects of a juxtaposition of transformation rules and tolerance clauses that do not grant equal importance to all manufacturing processes. The objective of this study commissioned by DG Trade is to assess the feasibility of introducing a value criterion in rules of origin for textiles. The rules of origin are the small prints of trade regulation. They receive little attention but are essential allocating preferential treatment for imports. For textiles, since the phasing out of quantitative restrictions on the 31 December 2004, preferential tariffs is a major tool remaining in trade the European Union has for fostering development and/or regional integration. Trade in textiles and clothing is essential for some countries. For Laos, Cambodia, Sri Lanka, Bangladesh, Vietnam, Tunisia, Morocco and many others, clothing represents more than 50% of export income. The apparel industry is the main source of earnings and the main industrial employer. It is also the most important source of income for women. In some other countries, like Lesotho, Malawi and Mozambique, the clothing industry can offer the start of an industrialisation process. The textile and clothing industries are old industries. Regarding Europe, it is enshrined in manufacturing traditions established in a regal age for weaving and in the industrial revolution for spinning. Tailor-making has a vibrant history in central Europe. Paris, London and Milan are hubs of the global fashion sector. Textiles are characterized by a complex set of processes from spinning to twisting, twining, warping, beaming, shearing, sizing, singeing, de-sizing, scouring, bleaching, mercerizing and all other stages that add value, not only as words on the Scrabble board, but also in industrial reality. However, textiles change and with the advent of electro-spinning, bonding, tufting, coating, laminating, encapsulating, plasma and CO2 finishing, the nature and value of industrial processes change. The global map of textiles has also changed. Before 1964, only 25 countries in the world had a substantial textile and clothing industry. Output was concentrated for 70% in industrialized countries. In 2005, at the end of the phasing out of the Multifibre Agreement, the industry has spread to over 60 countries with a sizeable industry. Industrialized countries are still important but represent less than 30% of industrial output. A new international division of labour has been firmly established. 4

5 In a changing industrial reality rules of origin are enablers or disablers of trade. The current rules of origin in textiles are precise, in fact alleged as extremely complex especially through the combination of basic principles, secondary rules, additional value criteria for some products, various tolerance rules, a patchwork of cumulation zones and derogations and territoriality clauses. They are also differing in their details across preferential schemes and ridden with tolerance clauses and exceptions to the general principles. The complexity of the rules is a reflection of the complexity of the industry, but they are also a reflection of an industrial age dominated by mechanical processes and a limited globalization of production and trade. They are also complex to understand, implement and monitor in a fragmented industry dominated by SME s. As the European Commission stresses the current situation, which combines, in a number of cases, complex rules with weaknesses in their enforcement, is not satisfactory Com 2005/100 Final. As the European Commission communicates to the Council and the European Parliament, the objective of a reform of rules of origin is (Com 2005/100 Final): To make rules simpler and, where appropriate, more development-friendly, the Commission would support: A simplification of the concepts and methods used for the purpose of determining origin ( ); such simplification should improve clarity, aid comprehension of the rules and support their application and enforcement. An adjustment of the conditions imposed on production processes to confer originating status, insofar as development policy and developing countries are concerned, with a view to guaranteeing easier access to the Community market through preferential tariff treatment corresponding to the real production and export capacity of the beneficiary countries, in particular for the least developed and smallest countries; An additional relaxation of the conditions to apply cumulation of origin within coherent regional groupings, subject to appropriate mechanisms in place for administrative cooperation between the partners to cumulation. The European Commission does also announce that in a number of sectors including agricultural, fishery or textiles products, moving to a new method for the determination of origin will represent a major change whose impact needs to be properly evaluated in advance. Should this evaluation demonstrate that the value added approach would not deliver the expected results in terms of development and simplification for certain sectors; the Commission will adopt another approach to better achieve these objectives. Moreover, the value-added criterion may, where appropriate to prevent possible misapplication or circumvention of preferences, be supplemented with additional conditions or criteria supporting actual development cit. op.ibid. This report is the final report in the Study on the application of value criteria for textile products in preferential rules of origin (tender Trade 06/H13), being supportive to the impact assessment announced here above. It combines desk research and selected field research and brings forward conclusion from modelisation as on field research. The final research draws on a field case in the GSP, one ACP country and one FTA with a Mediterranean country. It also draws on an examination of impact on the EU industry. The assignment for the European Commission regarding the feasibility of value criteria in preferential rules of origin can be summarized to five main questions: 1. What is the present structure of rules of origin in textiles and what is the state of play on their limitations on trade and development? What are the alternatives that have been tested or are under discussion? What are current and past experiences with reform of rules of origin? 5

6 2. Is the introduction of such a criterion justified by trends in trade? Do current trends in textile trade provide justifications for changes in origin rules? These trends should in particular be considered against the background of full liberalisation of textile trade in quantitative terms in Is it possible to establish a (set of) value criteria that assists development and discriminates well between exports that do or do not contribute to development? Is such a value criterion universal to all preferential regimes and to all textile products or is it justified to have a modulation by level of development, by fibre type or by stage of processing? Is it conceivable to reinforce or combine value criteria with other tools in trade policy or custom law of a temporary or permanent nature? 4. Will a change in rules of origin assist smaller supplier countries and less developed countries in growing their market share in international trade and hence contribute to sustainable development? As far as origin rules are concerned, this could be measured with three criteria: Growing market share in EU textile imports. Upgrading through rising export value with growing volumes and higher added value per unit of product. Broader industrial base by covering a larger section of the pipe line, more particular by promoting backward integration. What is the impact of suggested value criteria on levels of trade? Does it have a significant positive impact on the countries and policy objectives desired, while considering impact on EU production? This both being based on modelization as on expected behaviour of economic agents, such as derived from existing research and selected field studies. 5. Can a set of value criteria be implemented and is it enforceable? It is indeed a simplification for economic agents and authorities, also taking into account introductions of concepts such as net production costs and registered exporters. The researchers are grateful to the supervision and comments from the European Commission, in particular DG Trade and DG Taxation and Custom Union. The researchers are further indebted to the many companies in the EU, Sri Lanka, Turkey, Tunisia, Lesotho, Mauritius that have provided data and insights for the research. All arguments and conclusions do nevertheless only engage the consultants and do in no way reflect opinions of the European Commission. Enschede, March 15 th

7 METHODOLOGY The report starts with a discussion based on the main principles of current preferential rules of origin and its policy framework (chapter 2). It does then present alternative models and brings both alternative models and the EU current rules in debate. The proposal to introduce a value criterion is also reviewed in the literature. The objective of this chapter is to have a definition of criteria and hypothesis to be examined in the report. In chapter 2 we will carry out an analysis of trends in trade in section IX of the HS based on Comext figures for the period We will further focus on chapters 52 (Cotton) and 55 (Staple fibres) as well as on clothing in chapters 61 and 62. The trend analysis distinguishes main groups of countries: Pan-Euro Med (e.g. Tunisia), ACP (e.g. Mauritius), GSP Middle Income (mainly ASEAN e.g. Thailand) and GSP Low Income (e.g. Bangladesh), as well as Latin America (e.g. Peru), Euro/Med/Non Accessants (e.g. Ukraine). It will also refine the classification of countries in take off, ceiling or stupor patterns The first have shown substantial increase in exports since 2000 (e.g. Bangladesh). The second group has had stagnating levels in exports (e.g. Sri Lanka). The third group has not developed any substantial trade level (e.g. Lesotho). The classification of countries allows for an analysis of the specific conditions required for take off and possible conditions leading to stagnation or stupor in textile trade. This can be due to political context and risk, economic context but also to position in regional trading blocs or in the global trading system. This analysis aims at contextualising the possible impact in changes in rules of origin. This provides input for the field research and for impact analysis. Changes in sourcing strategies will be also assessed. With deindustrialisation in Europe, there is a shift from subcontracting relations (Outward Processing Trade - OPT type transactions) to sourcing of full package (Gereffi, 2001). This means that suppliers provide more service and hence see an increase of added value outside of direct manufacturing costs. However, there is also a tendency of formation of multinational manufacturing groups, which develop foot-loose branch plants to profit from specific short term economic and political arrangements. Both trends may have an impact for value criteria, especially in connection to concepts of net production value in origin rules. This research stage is based on desk research and some interviews with key players in sourcing processes. In chapter 3, based on interviews, proxy calculations and trade data analysis, a first examination of value criteria will be done. This is still based on a limited sample of calculations, costing calculations collected in Sri Lanka and on a simulation model based on standard cost parameters for materials and making up costs. It should be said that there is a wide range of methods to reach calculations, by taking trade statistics, VAT data or data used for calculating GDP, company data and calculations on single products or product categories. It is extremely difficult to come to a good comparable basis as companies use different methods of calculation; moreover there are issues with confidentiality attached to supplying commercial data. Hence our data is still tentative. But the ultimate objective is to set up a database of cost prices for the most important products and product chains. A further refinement will be needed if a comprehensive modulation across products and production processes seems justified. This section is based on desk analysis and some key interviews with most relevant technology holders (e.g. textile machinery manufacturers). The conclusions from chapter 3 will enable to select a scenario for impact assessment. The work will be carried out with two value added scenarios. A relatively high value added scenario in the range of 70-90% and a relatively low value added scenario around 45-55%. This can be modulated according to level of development. The High Value criterion range replaces the double transformation criterion and attempts to simplify and 7

8 harmonise. A low value criterion range will be calculated in order to give LDC s conditions for take off. In addition, the scenario definition will include several scenarios for cumulation of origin including EU-bilateral cumulation and regional cumulation as well as cumulation between all preferential regimes. Simultaneously, it will determine how the rules regarding preferential origin may avoid that beneficiary countries will only be used as mere platforms for minimal processing of goods from non-beneficiary countries or from beneficiary countries that are excluded from preferential treatment (such as e.g. China or India). Also in this respect, the focus will obviously be on the sufficient processing thresholds (SPT). However, besides the SPT s, also the options with respect to cumulation and/or possible other safeguards, such as e.g. insufficient working will be evaluated. The econometric analysis will be based on a scenario approach. The essence of the scenario approach is to facilitate a decision process based on specific values and objectives. Especially in complex decision processes the experience is that it is hard to give insight into the impact of the various options on the objectives of the client. This approach enables us to make the various options comparable expressed in the objectives set. Assessing the effects for economic operators, both in developing countries and in the EU, of switching to a value added criterion in the rules of origin for the textiles and clothing sector is a complex task to be included in Chapter 4. Various methods need to be used in order to evaluate the different dimensions of the question. In this study we propose to use cost-benefit analysis as well as economic modelling techniques. The latter is particularly appropriate for taking into account the complex interactions in the economy generated by a change in the rules of origin. In a GCE, a modification in rules of origin is seen as an external choc wherein a modification of rules of origin affects the real level of import duties and thus relative prices. Modifications of rules have also a trade creation effect in mobilizing production capacities that are now not used. Besides examining the potential partial impacts of the change on the economic operators at the micro level, we need to go beyond the partial-equilibrium analysis and use a more elaborate framework which is able to take into account the general equilibrium interactions between operators, markets, and countries. General equilibrium modelling provides such a framework. This is why - besides the qualitative and quantitative partial equilibrium analysis and costbenefit analysis - we propose to use the dynamic general equilibrium modelling platform developed by the Global Economic Modelling Network. This modelling platform allows us to quantify the potential impacts of a change in the rules of origin on the textile and clothing industry, employment, trade, production, income in developing countries as well as in the EU. General equilibrium models are based on microeconomic theories. They are designed to measure the direct, indirect and induced economic and environmental impacts of policy changes on an economy in the short, medium and long run. The inputoutput core enables the model to trace the extent and the channels of changes in policy and international environment. The resulting price changes affect the demand for the sectoral outputs and alter the resource allocation of factors. General equilibrium (CGE) models simulate the workings of a market economy and are unique in their ability to analyze the impacts of economic policy decisions, especially when the policy has macro, and sectoral resource allocation repercussions. CGE models are superior to traditional macroeconomic models because macroeconomic models cannot capture the structural impacts of policy shocks. CGEs are explicitly designed to capture all structural impacts including changes in relative prices, demand composition, and sectoral output and employment. 8

9 The country analysis entails a desk analysis and a field research. The country field cases enable identification of bottlenecks, both in specifying value added criteria and elements of implementation. The country cases will be exemplary and must represent the current limitations of origin rules. If we adopt the distinction between take off, ceiling and stupor countries, we favour a focus on the latter two types. Especially the last group requires a qualitative analysis, as the level trade is too low to enable significant extrapolation effects in modelisation. According to the methodology proposed we have proposed Sri Lanka and Tunisia as countries that have reached a ceiling in their exports, while Lesotho is selected as a case with no substantial trade with the E.U. Lesotho has build up a clothing industry under a favourable AGOA scheme, partly linked to South African industry and partly with Asian capital. The final chapter deals with effects in terms of simplification, clarity, application and enforcement point of view. In the interim report we focus on effects of the value criterion itself, with specific reference to the proposal to depart from an ex-works value and introduce the concept of net production costs. The aforementioned also considering the proposed change in the establishment of the originating status, which would be done by the exporters themselves, instead of by the (independent) authorities of the concerning beneficiary country and the responsibility of the importer claiming the preference. How does an importer have to get certainty if the net production costs (NPC), thus excluding profit, are the basis for determining origin? In that respect, the importer has to rely on his supplier whilst he is responsible himself if the statement of that supplier on the originating status is wrong. The analysis will consider current experience with value criteria using current product cases as a proxy, supply chain tensions (transparency of calculations) relationship with economic value and implementation of control (issuing authorities, responsibility of the importer). The findings and opinions expressed are the clear and sole responsibility of the consultants and can in no way be attributed to the European Commission or engage the European Commission. Enschede, December 20th Prof. Dr. Michiel R. Scheffer 9

10 1.0 PREFERENTIAL TRADE AND RULES OF ORIGIN Rules of origin are rules aimed at establishing the economic nationality of goods. In non-preferential rules it is always possible to trace the nationality of a good as they adhere to a principle of wholly obtained or of a sufficient process representing a substantial transformation. Preferential rules of origin are discriminatory by principle as they accord a preferential status if conditions are met that are different from non-preferential rules of origin. A good does always have a non-preferential origin but does not always have a preferential origin. In preferential trade these rules are designed for determining whether goods are eligible for preferential treatment. Preferential rules of origin can be considered simple technical conditions in order to determine origin. There is, however, a wide consensus that rules of origin are more than that. The Commission states in the Green Paper that Preferential origin rules need to be fundamentally reviewed, especially in view of the level of duties likely to emerge from the new round of multilateral trade negotiations, the role to be played by preferential origin rules in free trade agreements and the policy of market access and supporting sustainable development COM (2003) 787 final. In the report to the European Parliament, the Commission explicitly states that rules of origin may hamper use of preferential access in woven clothing, and that a review of rules of origin is required. In most recent communications, the Commission has acknowledged that preferential rules of origin have an industrial and a developmental impact. In its Communication to the Council and the European Parliament, the Commission has stated to favour using a method of evaluation of sufficient processing based on a 'value added test' as the starting point. Under this method, a product resulting from the working or processing of imported non-originating materials would be considered as originating if the value added in the country (or in a region in the event of cumulation), amounts at least to a certain threshold (a minimum 'Local or Regional Value Content'), expressed as a percentage of the net production cost of the final product COM 2005/100. It does also assert that in a number of sectors including ( ) textiles products, moving to a new method for the determination of origin will represent a major change whose impact needs to be properly evaluated in advance. Should this evaluation demonstrate that the value added approach would not deliver the expected results in terms of development and simplification for certain sectors; the Commission will adopt another approach to better achieve these objectives. Moreover, the value-added criterion may, where appropriate to prevent possible misapplication or circumvention of preferences, be supplemented with additional conditions or criteria supporting actual development. The communication clearly stresses (page 9, COM 2005(100) that three sensitive sectors deserve a particular attention, i.e. fisheries, agriculture and textiles). As a matter of fact, the decision to launch this study is related to this paragraph, as this study has the objective to provide a more precise assessment of the impact of a reform in terms of feasibility and impact on both developing countries and on member states. In order to understand the implications of such a modification this chapter will in general terms deal with the different preferential trade regimes and their subsequent treatment in this report in section I. In section II we will describe the main principles of preferential origin rules as they currently exist. In section III we will examine the debate on origin rules in terms of impact and experiences 10

11 with modification in rules of origin. In the entire chapter examples will be drawn from textiles, as textiles are the subject of the research. Textiles are in general described as all products covered by chapters 50 till 63 of the Harmonized System. Textiles in the strict sense are covered by chapters 50 till 60, while chapters 61 till 63 are in fact apparel and other made up goods. 1.1 Preferential Trade and Rules of Origin Preferential Trade is an important feature of the EU s trade policy and has particular relevance for textile products. In % of textile imports came from countries with the possibility of full duty relief under FTA s or GSP. A further 25% of imports originated from countries with a possibility of partial duty rebate under the GSP. To be more precise only 63% of eligible textile imports was granted preferential access to the E.U. The difference between both figures Hence only 63% of the 62% eligible import streams are accorded preferential treatment, i.e. 39%. It is precisely that not all trade between preferential partners fulfils the criteria of processing in order to confer preferential origin. In general countries in the Euro Mediterranean zone achieve a high rate of utilisation of preferential regimes (between 70 and 90%) because of bilateral cumulation with EU Origin. The rates of utilisation under the GSP system is, however much lower, typically around 50%. For many countries (in particular ACP and Euromed) it is largely linked to the fact that a majority of GSP beneficiaries utilize other preferential channels and then reach a higher utilisation rate of preferences. However many Asian and Latin American countries do only have the GSP system and are possibly more contrained by the rules of origin in their exports. This figure is lower for Chapter 62 and higher for Chapter 61 while in both chapters there are substantial trade flows. The figure is also higher for other textile chapters but with lower and declining trade level. It is generally acknowledged that in woven products rules of origin are considered as having a limitative effect on exports of LDC s. A more precise examination shows that the use of preferences results from a combination of preferential margin and eligibility of products. This analysis is carried out in chapter 2. Relevance and Regimes Preferential regimes and hence tariffs are justified under two headings. In the first place all is in the framework of regional integration and in the second place as a tool to promote development. The GSP system has been reviewed every ten years. In terms of FTA s one may distinguish two generations of FTA s. The first in the 1970 s were intended to foster development in partner countries, and not leading to accession, and were asymmetric in benefits. The second generation agreements of the first half of the 1990 s were mainly aimed at preparing for accession and had symmetric benefits but with asymmetry in periods of transition. Technically speaking, the EU has two systems to grant preferences: 1. FTA or similar bilateral agreement (Cotonou Agreement with the ACP) 2. General System of Preferences, autonomous as well as Western Balkan and overseas territories) The FTA s confer full suspension of duties for products with preferential origin that are covered by the agreement. FTA s are relevant for most Mediterranean countries, and countries in accession to the E.U. (+ Mexico, South Africa, Chile) The Cotonou Agreement can be assimilated with an FTA although it is technically 11

12 not a country by country agreement. Cotonou is currently to be succeeded by EPA s: Economic Partnership Agreements. The GSP is an autonomous and unilateral system based on a modulated tariff reduction according to sensitivity of products. The GSP system confers a reduction of duties for textiles by 20% for many countries and a full suspension of duties for least developed countries (EBA) and countries benefiting form the so-called "GSP+" regime. The GSP system entailed from 1997 onwards an environmental clause and a labour clause providing extra reductions of duties as well as an exclusion of preferences either for environmental/labour reasons or an exclusion as the country has a dominating position in a category of products ("graduation"). Under the "GSP+", replacing as from July 2005 the two labour/environmental clauses on one hand and the "Drug regime" on the other, the full suspension of duties has been enlarged to a more generic good governance clause, covering ratification of ILO, some UN and environmental conventions, including a specific clause for beneficiary countries combating drugs trafficking. Since the GSP1995 there is an exclusion (graduation) clause for some sectors and countries. As far as textile and clothing is concerned and since 2006, this clause applies only to China for textiles and clothing and India for textiles. The graduation is justified on base of their dominant share in the GSP imports to the E.U. Singapore, Hong Kong and South Korea have been excluded from the GSP as their level of development did not justify preferences since Five categories of countries fall under preferential systems: 1. Developed countries with a symmetric preferential treatment such as Switzerland and the EFTA countries with full duty relief. These countries are not examined in this study although conclusions of this study could be relevant to them. 2. Pre-accession countries (Bulgaria, Romania) with symmetric preferential treatment and full duty relief. These countries are not examined in this study although conclusions of this study could be indirectly relevant to them. 3. Countries in transition (in Europe e.g. Moldova) as well as Mediterranean countries (e.g. Tunisia) and newly industrialised countries with FTA s or being part of the GSP system with full duty relief. 4. Newly industrialised and lower and mid income countries with FTA s (Sth Africa) GSP or GSP+ Policy and full duty relief (e.g. Sri Lanka). 5. Least developed countries under the GSP system (EBA) or under the ACP Convention (e.g. Lesotho, Bangladesh). In textile terms the preferential regimes and for the scope of this study we will independently of the detailed preferential rules classify three groups of countries: 1. Developing and transition countries in the Pan-Euro-Mediterranean Area with either a FTA or GSP based preference a full duty exemption and good use of bilateral cumulation. Example: Tunisia 2. Developing countries in Asia and Latin America with either a FTA or GSP based 20% reduction on MFN tariff, limited use of bilateral cumulation and some use of regional cumulation: Example Vietnam. 3. Developing and LDC countries in Asia, Africa, Latin America, Caribbean and Pacific under EBA, GSP+ and ACP systems with full duty exemption, limited use of bilateral cumulation and some use of regional cumulation. Example Lesotho, Sri Lanka. To summarize one examines broadly two regimes (GSP vs. FTA), two levels of development (developing and LDC), and two geographical ensembles (pan- 12

13 euro/med and other). The level of development coincides with competitiveness in cost terms; the geographical ensemble coincides with time constraints and complementarities in industrial competence. 1.2 Architecture of Preferential Rules of Origin As far as current rules of origin are concerned the following elements come into play as having effect under current or reviewed preferential rules of origin: Sufficient Working or Processing Cumulation (and value criteria within cumulation) Tolerance Rules Territoriality Absorption Sufficient Working or Processing Preferential rules of rules of origin for textiles can be based on any of the following methods and combinations thereof: Wholly obtained Change of Tariff heading Value Criterion Specific operations to be mentioned in the so called single list. Wholly obtained is usually not relevant for textile products unless they are obtained from animals (e.g. wool), plants (e.g. cotton), mineral resources (e.g. synthetic fibers) provided that the resource are themselves wholly obtained in the country of origin and is further processed. This is the clearest case for ginned cotton and cotton processed into slivers or yarns, e.g. originating from Western Africa. As far as textiles are concerned, change of tariff heading is uncommon in chapters 50 till 63. For almost all products specific operations are mentioned in the single list. The overarching logic is that of double transformation. Double transformation is not a term in a legal or formal sense but the concept used to amalgamate the requirement of two substantial transformations. The most used transformations defined are the transformation of raw materials into yarn, yarn into fabrics and fabrics into assembled products. In general a single transformation coincides with a change in tariff heading. For most products two transformations are required. The most notable exception is yarn, where single transformation (spinning) suffices provided that preparatory operations (combing and or carding) are performed. For products that are obtained as end products by simple operations or by intrinsically combined operations a triple transformation is required. This is the case for simple made up goods such as bags (where the test is manufacturing from combed or carded fibre) and socks (where the knitting process itself forms the end product and manufacturing from fibres is required). For a limited range of products a simple transformation (not implying a change in tariff heading) suffices but the operations to be carried are specified (e.g. printing together with two operations) and complemented by a value criterion (generally 52,5% value added). This relevant for printed fabrics but also for coated fabrics and some technical products in chapters as well as for embroided products in chapters 61 and 62. Thus value criteria do already exist in textiles and clothing. 13

14 It is also important to assess that it is mainly the mechanical processes that participate in determination of origin, which is spinning, weaving, printing, coating, assembly and embroidery. Chemical processes, such as scouring, bleaching, washing, dyeing, finishing, are only considered as additional elements, not as participating as key elements in determining origin. One mechanical process is not considered as substantial processes as far as preferential ROO are concerned: cutting of fabric. It is also memorable that some non manufacturing processes are not taken into account such as patterning, grading and marker making that are important preparatory steps and cost elements. It is important to keep in mind for further discussion as some of these processes add substantial value. To summarize, a value criterion has now limited significance in textiles, as it is only a supplementary condition Cumulation Cumulation allows carrying out several substantial transformations in different countries/territories being part of an agreement. Cumulation is only allowed if the product in the last country of transformation has undergone substantial processing (e.g. clothing assembly) and in most but not all preferential agreements if the used material has itself already obtained preferential origin (the exception is the so-called full cumulation applying for ACP). That means that a garment made of a fabric that is of preferential origin receives preferential status if the country of assembly and the country of fabric are both part of a cumulation zone. The cumulation rule requires thus a triple transformation in the preferential zone, unless the intermediate product qualifies for preferential origin because of an alternative origin criterion (e.g. printed fabrics). There are exceptions to the triple rule in the first generation of preferential agreements, signed between the EU and third countries in the 1970 s where the final product is granted preferential status when two transformations have occurred in the preferential zone. The EU has different types of cumulation in its FTA and autonomous systems: bilateral cumulation, diagonal cumulation (and a similar provision in the GSP, i.e regional cumulation) and full cumulation. Bilateral cumulation is between the EU member states and the third country (also called donor content), diagonal cumulation is between third countries but almost always within a regional context e.g. a regional integration scheme. While bilateral, diagonal and regional cumulation require that minimum processing occurs in each of the countries (cumulation of already originating products), in full cumulation all processes are allowed as long as the final product meets the double transformation rule in the cumulation zone (cumulation of working and processing operations). A value criterion is introduced in cumulation rules in the GSP system for ASEAN, SAARC and Andean/CACM. The example most referred to in this study is the possibility for Sri Lanka and Bangladesh to use e.g. Indian materials (having preferential status) in end products as long as the value of the materials is less than the added value of the end product. This more or less approximates a 50% value criterion. Regional cumulation does only enter into effect when and if countries in the regions have set in place similar cumulation rules and implemented the measures to comply with this provision. The existing cases provide for a good proxy of value criteria in preferential rules of origin. The only de facto case working is Sri Lanka in relation to the SAARC. Article 76 of Reg. 2454/93 envisages derogation from some origin requirements. The derogation enlarges the scope for regional cumulation and puts no threshold 14

15 on added value in the last country of processing as long as more than minimal operations are carried out. This derogation is restricted to LDC s that apply for it, and is supposed to be limited in scope and quantity. It is also supposed to be limited in time. The derogation seems designed to derogate for a narrow band of products in relation to infant industries, in order to protect existing investments and planned investments. The derogation is in fact limited to the value criterion within cumulation rules, but is limited to the use of SAARC, ASEAN or ACP-EU materials. Requests to allow for the use of Chinese fabrics have been always been rejected. However as the derogation has been granted to Cambodia and Nepal in 1997 and has been extended several times. It should shortly be prolonged pending the current/on-going review of the rules of origin Tolerance Rules Tolerance rules allow to use non-originating materials to a quantified extent. The quantity is depending on the type and age of agreement specified in weight or in value at a level of 8% or 10%. The tolerance rule makes a distinction between non-textile inputs (such as metal buttons) that can be used without restrictions, non textile inputs containing textiles that are also considered as non textile items and can be freely used (e.g. zippers), other textile elements that can be used up to a maximum of the tolerance rate (e.g. sewing thread, textile labels, and small parts in the outer shell, and linings and interlinings that are excluded from the tolerance rule. Tolerance rules should be concerned with substantial attention in a modification of rules of origin implying the introduction of a value criterion. In the first place the exclusion of linings and interlinings creates a substantial additional barrier for countries with a limited textile industry and hence creates an extremely high marginal tariff barrier to the benefit of EU lining and interlining manufacturers, especially for those countries with 0% duty access. The definition lining is sometimes vague and pocketing or belt fabrics van either be considered as a type of lining or as a part of the tolerance clause. In the second place in some products the part of non-originating materials may be substantially higher than 10% in the case of complex garments with a wide range of inputs. In the case of brassieres polyamide fabrics may form the majority of the materials used in weight, but a minority of lace in weight but representing more than 10% of the value of the garment may lead to a loss of preferential status for the good. Another well documented case is the use of high visibility tape in security clothing. While the basic material is often of EU origin, and making up is carried out in the pan-euro Mediterranean zone, the tape which is in fact a coated textile is of US origin and represents more than 10% of the value of the garment. Thus the substantial use of a non-originating element cancels the use of preferential status Territorial Continuity The condition of territorial continuity implies that all subsequent stages of manufacturing must be carried out without leaving preferential countries. Goods must be carried directly (directly consigned) from the last country of processing to the EU and may not undergo manufacturing processes in any country between two stages carried out in preferential countries. Only unloading, stocking and loading in a third country is allowed but under custom supervision. The main reason of this clause is to avoid transhipments and to ease control of the use of preferences but has an additional industrial impact in textiles. 15

16 The effect of the clause of territorial continuity is, inter alia, that woven materials of preferential origin are not allowed to be dyed or finished in a third country before being assembled in a preferential country, when qualifying for reduced rates. This means that while the preferential rules of origin are content with weaving and making up in a country with preferential status, the territorial continuity implies that dyeing and finishing (a non substantial process) needs also to be carried in the country, or in case of full cumulation in another country having relevant preferential status. This clause is both relevant to the current rules as it introduces an additional stage of processing that requires its specific machinery, skills and natural endowments as well as for future systems of determining origin as ceteris paribus a value criterion is to be based on subsequent and territoriality connected processing, if this clause remains in place. For the sake of the argument, almost all textile products require finishing operations after weaving. The clause of territorial continuity does not apply to the regional groupings identified above nor within the Pan-Euro Mediterranean Zone Absorption Principle The absorption principle states that non-originating materials that have gone through substantial processing are considered as having become originating in further processing. The absorption principle means that in considering the value criteria, the current principle is that the value of non-originating materials is not accounted for in the value of a product if it has been substantially transformed. This means if cotton is spun and woven, the non-originating fibre can be considered as originating. The precise definition of the absorption clause is an important one when it comes to using a value criterion to determine origin. This also means for a fabric made from a blended yarn that if a polyamide elastomeric yarn has been made with originating polyamide mantle representing more than 50% of the value of the fabric, around a non-originating elastomeric core representing less than 10%, of the value of the fabric, the resulting knitted fabric will be considered as entirely originating. 1.3 Discussions on Reforms of Rules For a long period of time discussions on rules of origin where confined to the margins of policy making and academia. They were limited to considerations of investment and local content and custom procedures. In its Green Paper (Com (2003) 787 final of 18/12/2003) the European Commission has acknowledged that the current rules do present limitations to LDCs to fully use their preferences. They do not any longer reflect the industrial reality in the European Union and they are complex to understand and implement. It is also acknowledged by the Commission that the level of complexity renders trade facilitation complex and the quality of control does often not match the complexity of the rules. The debate on rules of origin is confined to a small group of academic scholars, but the body of research is increasing alongside the growing policy interest for rules of origin as trade policy tool. The interest comes from three disciplines: international law and politics, with a focus on legal coherence, international economics and developmental economics with a focus on globalisation and 16

17 qualitative effects, econometrics with a strong focus on quantitative elements and modelisation of trade effects. Besides the academic debate, rules of origin are the subject of intense and constant attention from the industry. Euratex has a committee on preferential rules of origin since beginning 90 s. In countries like Bangladesh, South Africa rules of origin are subject of intense debate between textile and clothing industries. Interest from NGO s into rules of origin and textiles is more recent, but they have certainly joined the debate. Most notable is the working paper of Oxfam 1. We will first address in this chapter three cases of reform of origin rules in respect of preferential schemes. They provide valuable input for a wider debate that follows in this chapter. The USA has introduced in the AGOA act a temporary single transformation rule. In Canada the Market Access Initiative for LDC s entails for LDC duty free access with a low value added criterion (25%) and a GSP wide cumulation zone. In Australia a debate and policy shift is moving from value criteria to change of tariff heading, since development is not an argument for reform this case is not analysed in this report AGOA The AGOA system, fully named as African Growth and Opportunity Act, is a system of temporary custom duty relief granted by and for importation into the USA. It has three main features. It grants total duty relief, it has a double transformation origin rule with diagonal AGOA cumulation but has also and it is a temporary relief system of the double transformation for the least developed sub Saharan countries) in AGOA. This derogation is subjected to one quota allowing a total import from AGOA Countries to the USA equal to 4,74% of total apparel imports into the USA. The AGOA system has had a major impact for apparel industrialisation in countries like Madagascar and Lesotho. The export level developed by AGOA countries to the US is substantial, at a level of 1,2 billion for clothing while the exports to the EU are below 0,7 billion. If one excludes Mauritius the ration is 1,0 billion against 0,3 billion. Exports of Kenya, Lesotho and Swaziland to the USA are together more than 720 million, the level to the EU is below 10 Million. The AGOA system has indeed enabled substantial use of fabrics of South East Asian and Chinese origin: 85% of imports are based on non-origination fabrics. This figure is above 95% for all countries but Mauritius and South-Africa. It has also led Taiwanese, Hong Kong China and Korean groups to set up factories in countries such as Lesotho. Madagascar has had more of a Mauritian led industrialisation. The trade effect on US imports is however small. The main effect is a trade diversion from China to the AGOA countries and from the SAARC countries to the AGOA countries. As it will be addressed in the Lesotho Case, it is often assumed that AGOA has led to industrialisation with foreign ownership. The type of industrialisation can be qualified as neo-fordist, that is factories based on simple machines, a refined division of labour and hence the mobilisation of unskilled workers, trained to perform effectively simple operations under foreign supervision. These factories are sometimes considered as night shifting factories, easily installable and easily removable. In terms of trade effects the AGOA system has two impacts: a trade 1 Oxfam Stitched Up, Briefing Paper no.60 April