Report No. 34964-BD Bangladesh End of MFA Quotas Key Issues and Strategic Options for Bangladesh Readymade Garment Industry January 18. 2006 Poverty Reduction and Economic Management Sector Unit South Asia Region Document of the World Bank TABLE OF CONTENTS EXECUTIVE SUMMARY .......................................................................................................................... i I Introduction 1 I1 ....................................................... 2 I11 TheDomesticContext: Challengesto Overcome ... ................................................................................................................................. Post-MFAGlobalPerspective:Shapeof Thingsto Come ..................................................................... 7 I V Implicationsof QuotaAbolition ................................................................................................ 9 V 12 VI CompetitivenessandStrategicSourcingDecisions ................................................................ 18 VI1 StrategicOptionsfor BangladeshRMG ................................................................................. 19 VI11 Conclusions .....IndustryStructureand Sourcesof CompetitiveAdvantage ................................................. ................................................................................................................................ 52 REFERENCES .......................................................................................................................................... 55 ANNEXES .................................................................................................................................................. 59 STATISTICAL ANNEX ........................................................................................................................... 73 List of Tables Table 1: Export o f apparel and textiles to the U S and EU....................................................................... 3 Table 2: Economic welfare results from GTAP simulation..................................................................... 4 Table 3: Changes inUnskilled Labor after Quota Abolition ................................................................... 5 Table 4: Export o f RMG2001-05 ............................................................................................................ 5 Table 5: U S Import o f MFA Apparel. January . 2005..................................................................... July 6 Table 6: Summary o f anticipated effects o f quota elimination inU S market in2005 and key competitive factors. by selected countries.................................................................. 10 Table 7: 12 Firmlevel export concentration............................................................................................... Structure o f RMGindustry...................................................................................................... Table 8: 13 Table 9: Labour and power cost inselected countries ........................................................................... 14 Table 10: Estimated cost o f investment ................................................................................................... 14 Table 11: Summary Features o f Surveyed FirmsinRMGIndustries...................................................... 24 Table 12: Lead time for servicing an order for RMG.............................................................................. 27 Table 13: Typical leadtime components for Bangladesh........................................................................ 28 Table 14: 29 Relative prices o f typical knittingyarns .................................................................................. Consumptiono f woven and knit fabric by export-oriented RMG sector ................................ Table 15: 31 Table 16: Domestic production o fyarn and fabric .................................................................................. 38 Table 17: Illustrationo fprice advantage due to GSP scheme ................................................................. 39 Table 18: Changes inBangladesh Export................................................................................................ 40 Table 19: Welfare changes due to elimination o f tariffs for Bangladesh's exports of textiles 40 Decomposition o f Welfare Changes ........................................................................................ and apparel to U S..................................................................................................................... Table 20: 41 Table 21: India andBangladesh: Total Garment Exports 1997-2004...................................................... 43 Table 22: India. Mens' or boys' woven cotton ......................................................................................... 44 Table 23: Comparisons o f labour costs and productivity inthe RMG industry...................................... 45 Table 24: Prices of some Bangladesh exported RMGs and preferential tariffs for Bangladesh importsinIndia........................................................................................................................ 46 List of Boxes Box 1: Post-MFA Action Programme ................................................................................................... 8 Box 2: IndustryStructure andRMGExports...................................................................................... 13 Box 3: Comparative advantage: Basics versus highvalue added apparel........................................... 15 Box 4: Key Recommendations o fthe Growth and Competitiveness Report ...................................... 21 List of Figures Figure 1: Distribution o f firm productivity indifferent sub-industries .................................................. .23 Figure 2: Distribution of firmproductivity by location .......................................................................... 23 Figure 3: Productivity Growth o fDomestic Firmsby Sub-industries .................................................... 23 Figure 4: Productivity o f Firmswith Different Ownerships .................................................................. .24 Figure 5: Number o f FirmsinDifferent Markets.................................................................................... 24 Figure 6: Number o f Firms vs. Number o f Markets................................................................................ 25 Figure 7: Market ChoicebyFirmswith Different Markets .................................................................... 25 Figure 8: Exporting and Productivity...................................................................................................... 26 This report was preparedby a team led by Zaidi Sattar (SASPR). Other members o f the core team included: M. A. Taslim (consultant); Vlad Malone, Hiau Looi Kee (DECRG-Trade); Nilufar Ahmad (SASSD); Shafiqul Islam (MOC); Nasiruddin Ahmed (MOF); Rajani Alexander (CIDA), Frank Matsaert (DFID), Zillul Hye Razi (EU Commission), and F.M. Ziaul Ahsan (consultant). The team benefited from the expert guidance and contributions from Shanta Devarajan, Will Martin, Tercan Baysan and Syed Mahmood. The report was prepared under the general guidance o f Sadiq Ahmed, Sector Director; Ijaz Nabi, Sector Manager; and Christine Wallich, Country Director. It benefited from comments o f peer reviewers: Zalur Ahmed Khan, Syed Mahmood, Montford Mlachila, Zareen Naqvi, Philip Schuler, Paul Brenton, Peter Walkenhorst, Martha Pierola, and Mombert Hoppe. Aneeka Rahman and Nermeen Shams Rouf prepared the Statistical Appendix and Annexes and provided invaluable research support. Mehar Akhter Khan processed the report at its various stages with support from Joyce Mormita Das. The following individuals shared their knowledge and vision o f the textile and RMG sector in the course o f many discussions: M.A. Awal, Shabbir Ahmed, A. Matin Chowdhry, Ghulam Faruque, Annisul Huq, Fazlul Haque, Anisul Islam Mahmud, Wahiduddin Mahmud, Zulfiqar Rana, and Ramzul Siraj. This report was partly supported with funds made available by U K ' s Department for InternationalDevelopment (DFID) and Canadian International Development Agency (CIDA). Endof MFA Quotas: Key Issues and Strategic Optionsfor BangladeshReadymadeGarmentIndustry Executive Summary I.ThePost-MFAChallenge 1. "With the removal o f MFA quotas, there will be opportunities to expand market shares, but also tough competition for markets. The challenge i s not to protect a special position, but to open upthe marketsandensurethat Bangladeshwill beina strongpositionto compete."' 2. This study addresses that challenge. It examines thefactors that have brought Bangladesh such impressive success as aproducer and exporter of ready-made garments (MG) in a quota-based trading regime. I t explores the likely threats to that success as the post-MFA trade regime becomes intensely competitive. And it sets out a number of strategic options for the sector to pursue, building on past achievements and inherent competitive advantages, in order to compete successfully in textile and apparel markets under new and very demanding conditions. 3. The study suggests a dual approach to the challenge: a strategy to minimize the vulnerability that export concentration in readymade garments (RMG) creates for Bangladesh, and, secondly, a focused strategy for strengthening the competitiveness o f the RMG industry. To address the first challenge o f export diversification, the report takes on board the set o f recommendations made in the World Bank's Growth and Export Competitiveness Study (World Bank 2005a) which argued the case for minimizing the remaining anti-export bias o f Bangladesh's trade policy and removing behind-the-border constraints - including infrastructure weaknesses -- to overall export competitiveness. To address the post-MFA competitiveness challenge for the RMG industry, this study lays out a number o f strategic options requiring, inter alia,policy reform, institutional change, and infrastructural investments. 4. The study finds that reducing lead time - turnaround time from receipt of orders to shipment o f goods to market - for the woven garment sub-sector, ifnot for all RMG, i s critical for survival. Success in global apparel trade i s increasingly hingingon exporters' ability to react quickly to orders, to cut the lead time required to produce and ship apparel. To meet the competition from countries that either make their own grey fabric - a basic material intextile manufacturing -- or can import sufficient quantities on short notice, Bangladesh very much needs to raise RMG productivity, both through skills training and technology investments, especially foreign ones. FDI can be and regularly has been a powerful force behind improving productivity. Having only recently taken the first step toward easing the flow o f FDIby dropping restrictions too long inplace, Bangladeshneeds actively to court foreign investment. 5. Inresponse to the risingemphasis onrapidresponse, Bangladeshalso needs to improve its ability to quickly access supplies o fraw materials. Inaddition to urgent upgrading o f Chittagong port facilities, one option i s to establish Central Bonded Warehouse facilities in the Export Processing Zones (EPZ). Another - a policy choice too long deferred - i s to lift the ban on importing yarn from India by land routes. Importation by sea from India adds to transport costs and often takes 10-20 days extra to reach Chittagong port. The prohibition hurtsRMGindustries by undermining their competitiveness. Inthe post- MFA era, the casefor rescindingit appears compelling. ' "The Investment Climate, Governance, and Inclusion inBangladesh," Nicholas H. Stern, Chief Economist and Senior Vice President, The World Bank, speechto BangladeshEconomic Association, 8 January 2002, p. 13 i 6. Within the South Asia region, Bangladesh can gain ground from a changing external policy environment by seizing opportunities for preferential market access, taking advantage o f regional cumulation, and, if possible, negotiating free-trade agreements with India and other neighbors (SAFTA, BIMSTEC). At home, the government needs to work on buildingpartnerships with the RMG industry to attract FDI and improve marketing abroad by burnishing the image o f Bangladesh-made apparel and its association with leading RMG brand names. Recognizing that tougher competition and higher productivity will mean job losses in the sector, the government will also need to actively explore mitigating options for displaced garment workers. Whathappened since January 2005? 7. Model predictions versus actual performance. MFA quotas were phased out at the end o f December, 2004. The GTAP model used in this report predicted export shocks and economic welfare losses for Bangladesh, albeit with a lag. Over three-to-four years, the model predicted, the economic welfare costs could run to $370 million. Interms of jobs, Bangladesh losses could be significant - 17 percent o f the current workforce in apparel and 5 percent in textiles. India and perhaps Palustan could achieve notable gains in South Asia, but Chinese workers appear to be the major winners as demand for textile and apparel, including those o f their linkage industries, rises sharply with opening up o f global markets. To take advantage of a quota-less trading regime, China, it i s feared, could develop the entire range o f textile industries from cotton to ready-made garments. The surge inChina's import of textile and clothing related machinery appears to corroborate this concern. Between 2000 and 2003, China's import of textile and clothing machinery increased from about $2 billion to more than $5 billion. Some experts are predictingthat China may capture as much as 50 per cent o f the global export market. Table 1.Export of RMG2001-05 (US$ million) woven Knit RMG garments garments Total 2000-01 3,364.2 1,496.4 4,860.6 2001-02 3,124.6 1,459.2 4,583.8 2002-03 3,258.3 1,653.8 4,912.1 2003-04 3,538.1 2,148.0 5,686.1 2004-05 3,598.2 2,819.5 6,417.7 Year-on-year growth (9'0) 2001-02 -7.1 -2.5 -5.7 2002-03 4.3 13.3 7.2 2003-04 8.6 29.9 15.8 2004-05 1.7 31.3 12.9 Source: ExportPromotionBureau 8. The immediate post-MFA period has not revealed this trend. RMG exports from Bangladesh in the post-MFA world did not decline; rather, data show that export growthduring January-July 2005 was substantial to end the fiscal year 2004-05 with a growth o f 13% (Table 4), with knit garments registering phenomenal growth o f 31%. Indeed, performance inthe hitherto quota-restricted market ofU S A revealed inherent strength of the Bangladesh industryrelative to its competition (Table 5). During Jan-Jul, 2005, Bangladesh was clearly a winner with China, India and Sri Lanka, while countries like Hong Kong, Taiwan and Malaysia lost out. Knitwear exports to U S jumped 70% during Jan-Jun 2005 and BKMEA reports that order books are full. The rather pessimistic projections o f the model perhaps stem from the assumptions about quota premia. Clearly, quotas were more bindingfor Bangladesh than was presumed. .. 11 9. Yet, this should give no reason for complacency, as major challenges lie ahead: safeguards on China will be phased out by 2008; Indian textile and garment sector has undergone major technology uplift and policy reform to put India amongst leading contenders in garment exports; and inherent weaknesses in policy continue to plague the woven garment sub-sector, whose share in exports i s 56%, though declining. 11. RMGinBangladesh: PastAchievementsno GuaranteeofFutureMarket 10. Success in the coming contest for trade i s vital to Bangladesh's economic growth prospects and its hopes o f meeting the MDG goal of halving extreme poverty by 2015. As global RMG competition rises, Bangladesh cannot afford to be sidelined. The economy-wide effect o f losing ground could be traumatic. Three-quarters o f all 2003/04 exports -- $5.7 billion o f a total $7.6 billion - came from sales o f knitwear and woven goods, and 96 percent o f all RMG exports went to buyers in the U S and EU. Although the apparel sector contributes 5% to the country's GDP, it i s the biggest source o f industrial jobs. Along with accessory industries it provide direct employment for over 2 million workers, mostly women. Approximately eight million more Bangladeshis rely indirectly for their livelihood on the RMG industry. RMG not only accounts for 40% o f industrial employment, but has also become the main source o f foreign exchange earnings needed to finance vital imports o f capital goods and essential inputs while serving, as well, as an important training ground for a growing number o f new entrepreneurs. 11. A s the MFA phase out takes hold and as countries like China-India build ever more formidable RMG export industries, a substantial part o f Bangladesh's RMG workforce could be at risk o fjob loss if the industry fails to stay competitive. The global impact o f two measures - abolishing apparel quotas, and instituting safeguards in U S and EU against cheaper imports - have been explored through GTAP general equilibrium models. According to a simulation done for this study, global gains from quota abolition -- interms o f money-metric measures o f economic welfare -- aggregates to over $16 billion. The principal beneficiaries would be the U S ($6 billion), EU25 (9.4 billion) and China (over $3 billion.). This study presents a set of strategic options that begin, appropriately, with proposed measures to meet a crucial test o f competitiveness: reduced lead-time infilling RMGorders. 111. GearingUpto Compete:StrategicPolicyOptions 12. It is not a foregone conclusion that Bangladesh must fall behind in the race to supply global markets - especially those of the U S and EU - with ready-made garments and other textile products. That race, like others, will "be to the swift," and Bangladesh can take steps now that will accelerate its responsiveness and lighten the loadits RMG firms carry inthe contest. 13. Inthe post-MFA world, leadtime has emerged as a crucial determinant o fcompetitiveness. More than in the past, vendors need to show apparel trade buyers that they can fill orders fast. While several developments in the retail business have sharply reduced lead time in the U S and Europe, Bangladesh woven RMGmanufacturers have continued to lag. Among competing countries, their lead time i s one o f the longest (90- 120 days), principally because of inadequate, easily available supplies of local woven fabric and the consequent need to procure the raw materials from overseas. Ifthe raw materials could be sourced locally the time required to receive them at the factory gate could be reducedfrom 42-60 days to just one or two weeks, puttingBangladesh garment exporters on a par -interms o flead time -with their major competitors. ... 111 14. Improving theDomestic and Regional Supply Chain. The challenge is easier to describe than to overcome. The best solution - increasing domestic supplies - is not attainable in a short period, but - more easily and rapidly -- part o f the demand might be met by increasing the capacity o f dyeing and finishingindustries. Ifthey were allowed to stock upgrey fabric inadvance o f orders, they could supply finished fabric to the woven garment manufacturers at fairly short notice. Since the lead time would not be greater than when domestic fabric were used, this option merits serious consideration. 15. It will, however, take at least a few years o f buildingup the dyeing and finishing sub-sector to close or dramatically narrow the demand-supply gap. The gestation time would be long; but since the next few years will be crucial for the health o f RMG woven industries o f Bangladesh, other innovative tactics need to be explored. One could come from building closer relationships with nearby suppliers (e.g. in South Asia region) in order to receive fabric at short notice, a course without clearly foreseeable outcomes. Currently, although RMG manufacturers are free to import fabric from these countries, they have sourced only a small fraction o ftheir requirement from nearby countries such as India, Thailand and Pakistan. They may needthe incentive o f special advantages given them inorder to increase imports, but domestic textile manufacturers as well as competing countries are very likely to oppose such provisions vigorously. 16. Opening Land Routes and Modernizing Port and Customs Facilities. Even without new inducements, Bangladesh could markedly improve its utilization o f available means to reduce lead time by sourcing inputs from the regon, starting by rescinding the ban on importing yam from India by land routes. Whatever reluctance there may be to forging closer links with India or Pakistan for supply o f inputs at short notice has to be weighed against the urgent need to cut lead time by bringing yam to knitters and weavers as rapidly as possible. The policy now in place i s self-defeating. Among other things, its continuation would impede eventual free-trade negotiations with India or with other neighbors inSouthAsia. 17. The removal o f the ban should bejust a single element in a concerted effort to smooth the path o f imports - in some cases, smooth it literally by repairing the roads and modernizing the Chittagong port that constitute serious bottlenecks to the flow o f goods. It is important to remedy the power and transportation infrastructure failures that exporters see as heavy burdens on their competitive performance. While there i s no doubt that additional investment i s required in energy and transport infrastructure, the earlier World Bank (2005a) report concluded that the changes that are required most immediately are modifications ofpolicies, processes, and management rather than capital investment. 18. Inaddition to dealing with transport flaws, trade-related policy changes need to further simplify the import tax regime and reduce the dispersion and average level o f nominal (and thus effective) protection, preferably through a pre-announced medium- and long-term schedule o f tariff reductions (as done recently by India), and to eliminate any remaining protective quantitative restrictions. Specifically, to bring high payoffs in export competitiveness, reform should be a priority in Customs as well as the Duty Exemption and Drawback Office (DEDO). Practices in both these services are characterized by inefficiency and rent-seeking that penalizes firms capable, in a freer regulatory environment, o f expanding Bangladesh's share o f world trade. 19. Seizing Opportunitiesfor expanding market share and new markets. Positioning Bangladesh's RMG industry to compete successfully primarily involves changing domestic policies, practices and incentives. At the same time, Bangladesh should be seeking every opening available or in prospect to gain eased access for its products into U S and EU markets into countries like Canada, Australia and Norway which have granted duty-free access, and should even be actively exploring access to China and Japan, inaddition to opening up export opportunities within the South Asia region, such as inIndia. iv 20. Preferential access. As an LDC, Bangladesh has profited substantially from the EU grant o f duty-free access for its RMG exports, but still faces average tariffs o f 14percent on goods sold to the US. Not only does the U S not offer duty-free access to L D C exports, its tariffs are generally skewed against them to the point that Bangladesh paid more tariff fees on its apparel exports o f about $2 billion in2003 than Frenchexporters paid on their exports o fUS$26billion. By offering duty-free access to a number o f Caribbean countries and African LDCs, moreover, the U S has put Bangladesh exporters at a price disadvantage. Their response has been to lobby hard for the U S Congress to give them duty-free access under a proposed bipartisan bill called Tariff Relief Assistance for Developing Economies (TRADE) Act 2005. That bill would help several least developed countries - not just Bangladesh - and the work o f malung the case for wider access i s one that international organisations should be called on to support. 21. Were U S tariffs eliminated, Bangladesh apparel exports are projectedto increase by 90 percent in 3-4 years, and textile exports would grow by 82 percent. Gaining such a foothold and expanding market share inthe lucrative U S apparel market i s every garment exporter's dream, but Bangladesh needs to look at that prospect and preferentialaccess ingeneral with realism. It should take every possible advantage o f the preferences it can gain, but also understand the conditions attached and the reality that it will not remain an LDC forever. It should use the time and latitude to prepare for increased competition when it graduates out o f the LDC league. 22. Regional Cumulation. To gain EUpreferences, Bangladesh exports must meet a stringent test in the EUrules o f origin requiring that apparel be made o f domesticallyproduced fabric. Some Bangladesh products meet that test; others - made with imported materials --do not. While the EUreconsiders the application o f its strict rule to L D C exporters ingeneral, Bangladesh mightbe able, through the provision o f regional cumulation, to get around the bamer to preferential access. A beneficiary country, which i s part o f a regional trade area such as SAARC may claim, for GSP purposes, fabric purchased from regional member countries as its own when claiming GSP - as long as the value added crosses a high threshold -effectively over 50 percent for Bangladesh -o fthe finished product. Most Bangladeshi woven goods cannot meet that test and manufacturers are not likely to restructure their current supply chain to benefit from the narrow opening. Since EUrequires that respective Governments endorse the cumulation option, what i s needed i s for the Bangladesh Government to lobby for lower value addition criteria. Such a scheme offers incentives for sourcing inputs from the region, which, in and o f itself, could be a way for reducinglead time and enhancing competitiveness. Why pass up the opportunity? 23. Warehousingfor quick turnaround. Given the very large investment required and an adverse international market situation characterized by over-capacity and low prices, there i s little prospect o f improving the domestic supply situation in the short or medium term enough to fill the gap between supply and demand. An innovative approach-the establishment o f a centralbondedwarehouse (CBW) - can let local manufacturers match the lead time o f competing counties and do so quickly. A CBW differs from the individual bonded warehouse inthat each CBW i s permitted to stock duty-free imported inputs, while imports to it o f duty-free RMG and textile raw materials would not be conditioned on master export LCs. Inprinciple, the C B W operator could be permitted to stock up a whole range o f T&C inputs such as finished and grey fabric, accessories, dyes and chemicals, yarn, RMG and textile machinery and spare parts in amounts determined by expected demand. RMG and textile manufacturers can then purchase these inputsduty-free from the CBW directly as export orders are received. 24. Some appropriate system, as suggested by the National Coordination Council, will have to be put inplace (such as sale against back-to-back LCs and locating the CBW in Export Processing Zones) to ensure that the duty-fkee inputs are actually used in the production of goods that are exported, and there are safeguards against leakages to the domestic market. V 25. Because a CBW-system will make imported fabric available as quickly as, in some cases more quickly than, domestic fabric, RMGproducers who are procuring domestic fabric may switch to cheaper CBW supplies. This prospect generates opposition from textile manufacturers who want priority given to lead-time-reduction measures that would increase their production, not make competing imports more attractive. The reality, though, i s that Bangladesh's modem textile industry owes its growth to the RMG industry,and faster growth o fRMG-a goal dependent oncutting the time between ordered and delivered goods -- the larger will the market be for domestic textile producers. CBW offers a way to cut lead time, enhance competitiveness, and, thus to improve performance andprofit throughout the textile sector. 26. Textileprotection impedes RMG growth. It has also been suggested that the government should actively pursue policies that could quickly increase the capacity o f the primary textile sector (PTS) so that they could meet the entire demand for woven garments. To make this possible, the argument goes, existing measures that restrict fabric imports and thus protect the PTS should be maintained. Electing such a policy would run counter to the reality that no government can trulyjudge or accurately forecast the success o f a particular business. Governments do not and will not always choose winners. Indeed, they are more likely to bend to lobby pressure. The best policy for a government i s accordingly to take an even-handed approach to all industries by removing any constraints to fair market competition. The challenge for Bangladesh i s to adopt domestic policies that ensure a smooth supply o f raw materialdinputs to the apparel sector at international prices. Restricting supply inorder to protect primary processors i s unlikely to strengthen such manufacturers and certain to penalize RMGexporters. 27. Meeting New and Old Sourcing Criteria. Without the shelter o f MFA quotas, Bangladesh and its RMG competitors all face the same changed market conditions. Price still matters, and Bangladesh can retain an advantage in that area. Speed - lead time - matters more and more, and Bangladeshneeds to hurry to improve its performance. Quality consistency, reliability and flexibility, and service remain winning attributes, and Bangladesh has achieved and will have to maintain a strong record on those counts. 28. Other factors or combinations o f factors, however, have gained weight in the calculations o f developed-country buyers. Importers and retailers, for instance, are recasting their sourcing practices in order to better position themselves in a fast changing market. A 2004 study commissioned by the Commonwealth Secretariat on sourcing practices o f international buyers finds that major buyers o f apparel are increasingly adopting sourcing strategies that seek to minimize their total costs throughout the supply chain rather than minimizing the cost o fprocurement from the RMG factories at the lowest end o f the supply chain. This relentless drive for total-cost minimization by overseas buyers poses certain challenges for the RMG sector o f Bangladesh. In addition to price and lead time, these challenges relate to quality and to social and environmental compliance. 29. As part o f a strategic shift in global sourcing post-MFA, buyinghouses would like to limit the number o f countries and factories from which they source their apparels - without, though, putting all their eggs in the Chinese basket. In making country choices post-MFA, macroeconomic and political stability, infrastructure, compliance standards, and good partnerships, will be critical. Globally, therefore, many counties which held significant share o f apparel markets will face shrinkage or total elimination. According to various assessments, vulnerable countries include Nepal, Mauritius, Maldives, Sri Lanka, Thailand, Philippines, and Poland. Most analyses predict challenging times for Bangladesh, though all indicationsare that it will remain a major garment-exporting country. 30. Bangladesh gets mixed reviews on its compliance with many o f these criteria. On the positive side, buyers reveal that they are hardly ready to abandon suppliers with whom they built long-term relationships over the years and where they even invested in skill development, management practices and workplace conditions. At the same time, buyers who are now puttingextra emphasis on meeting ILO vi codes regarding worlung conditions and labor practices as well as adhering to international standards o f waste management perceive weak social and environmental compliance in Bangladesh. Infrastructure weaknesses, lack ofraw materials, and antiquated banking system are other problems cited by companies as industrydeficiencies that needto be overcome inthe days ahead. 3 1. Compliance with social and labor standards. As Bangladesh emerges as a major player in the global RMG export market, the pressure i s mounting for compliance with social and labor standards. Failure to conform to international norms could undermine competitiveness. It i s advisable, therefore, that RMG firms recognize these requirements as necessary for the health and safety o f employees and undertake that the standards set out in SA8000 are implemented and maintained. This i s important not only from the point o f view o f enhancing competitiveness and sustaining repeat orders, but i s also critical for keepingworkers motivated to improve productivity andproduct quality. 32. Bangladeshis mustforge a concertedpublic-private partnership to correct the weaknesses that interfere most with their competitive strengths. That partnership needs, among other challenges, to address serious shortcomings with regard to the country's international market orientation-the ability to gear up its marketing campaign in response to changing demand. Long years o f exclusive reliance on foreign buying houses or their agents have left ingrained habits that are out of step with the post-MFA demands for savvy merchandising and marketing o f products in order to penetrate and maintain market share in export markets. A partnership between government and the RMG sector must focus on such forward linkages. The following approaches highlight some o f the key strategies for promoting forward market integration: Product and market diversification: One o f the major limitations to apparel export growth is the lack o f diversification in products and markets - not least, as noted below - the giant market o f neighboring India. A small range o f products (shirts, trousers, T-shirts, sweaters, jackets) makes up 60 percent o f RMG export. Vigorous marketing and promotional drives would be needed to break into markets for other RMG products that Bangladeshi producers can produce competitively. Since R&D and market research to expand the product mix and public support in this area will be critical, a comprehensive research center built on public-private partnership should be established. Its role would be to gather and disseminate information effectively to local manufacturers on the latest developments in products and markets, including information on fabric developments, blends, colors, patterns, latest fashion trends and design forecasting, as well as providing customer service to foreign buyers purchasing fiom Bangladesh. Image building and brand development: Image building and branding can achieve a high level o f value addition and enhance Bangladesh's reputation as a quality supplier of apparel. As branding i s an expensive investment, incentives and opportunities should be provided to exporters, such as through a `Brand Fund' and by encouraging foreign collaboration to launch collective brand names through corporate marketing companies. Investments in skdls, design and advertising will be essential for the "Made in Bangladesh" label to make its mark in the global marketplace. For this, an active promotion campaign will be required by the industry, government, and Export Promotion Bureau. Local and international exhibitions should be held with aggressive efforts to attract foreign buyers. Bangladesh embassies abroad should appoint dynamic business-oriented commercial officers whose main job will be to promote export products through direct contact with potential buyers. At home, an exporters' performance rating database could also be developed, based on factors such as product quality, timely delivery and financial performance. This facility should be set up in collaboration with an accredited international company so as to ensure objectivity and credibility. vii 0 FDI or Joint ventures: Bangladesh has been a reluctant recipient o f FDI, at least inthe RMG sector, limiting it, until recently, to EPZ only while the domestic sector was kept off limits for foreign equity or partnerships. Less than 15 percent o f Bangladesh garment firms have foreign equity. This policy has had one positive impact - growth o f indigenous entrepreneurs who are now much more preparedto face the post-MFA challenge on their own. Although the restrictions on FDIwere finally removed inthe IndustrialPolicy 2005, the Bangladesh RMGsector may have missed out on certain benefits that accrue from the presence o f FDI in a sector. Our survey o f RMG firms showed those with foreign equity to have productivity levels on average 20 percent higher than those without. In addition, the survey showed positive and significant productivity spillovers. For every 10 percent increase in the productivity level o f FDI in the industry, productivity o f domestic firms increases by 1.4percent. Although Bangladesh has a limited number o fjoint venture RMG industries, even these can be very effective in bringing in managerial and marketing expertise. Continuous efforts should be made by the Export Promotion Bureau and Board o f Investment to ensure that the investment climate o f the future favors more FDI or joint ventures. The positive spillover effects o f FDI-- can make a substantial addition to Bangladesh's existing strengths. IV. Conclusion 33. The end of MFA presents both challenges and opportunities for the RMG sector of Bangladesh. There are many uncertainties about new market developments in the coming months. What i s certain i s that there will be heightened competitive pressure with the more efficient producers attempting to capture a larger share of the global market. To sustain and enhance Bangladesh's share, all stakeholders including the government, industry and individual firms will have to get behindpolicies and actions that help Bangladeshbecome more competitive. 34. Delay in adopting the right policies or removing harmful constraints can only raise the cost o f missed opportunities and reduce the competitive strength o f the sector that has made a remarkable contribution to industrial employment, woman empowerment, and reduction o f poverty. Above all, it would be a grave mistake to take any policy andor institutional action that might undermine the RMG sector's export competitiveness, and any existing trade policy distortion that i s harming this sector's competitiveness would needto be removed urgently. 35. The stakes are high.Trade competition i s mounting. Bangladesh has a strong position from which to compete and a wide choice o f promising policies to pursue. The one unacceptable policy choice i s inaction. ... Vlll V. Key strategic options and recommendations Kev strategic oDtions Recommendations 1.Raisingproductivity is key to Now that the FDI restriction has been removed competitiveness from the Industrial Policy 2005, all efforts should be made to facilitate FDI inflows into the sector to take advantage of better management, technoloa, capital and market access. 2. ReducingLeadTime i s critical. Overhaul Chittagong port facilities, improve infrastructure and transport logistics, and streamline the import regime by removing cumbersome procedures. End the ban on yarn imports through landports. * 3. RMG sector must not be held hostage to Ifthe Governmentdesires toprovide support to backward linkage. the textile industry, such as interest rate or price support, it must do so through its normal budgetary instruments in a transparent manner. The cost of support must not be shifted to the RMG industty that is totally unprotected - - in the internationalmarket. 4. CentralBonded Warehouse for quickening Permit the establishment of CBW in line with delivery time. the recommendation of the National Coordination Council. 5. Make the most of preferential market access Bangladesh should seize the opportunities which gives price advantage. offered to it as an LDC but use the space to preparefor increased competition when it graduates out of theLDC league. 6. Take advantage o fRegionalCumulation Government should lobby hard withE U to reduce value added requirement along with liberal regional cumulation option. 7. Compliance with social and labor standards To remain competitive and win orders, RMG ~~ ~ firms must conform to international standards of social compliancefor their industry * The Government has since removedthe ban o n y a m imports over landroutes for 100%export oriented industries. ix 4 3- X a 3E d .3 x ~ a i;E ! 7 E .fi t c 0 Le +L c F i PE I i 4I - Y U n v) m 0 U 0 Endof MFA Quotas: KeyIssues and StrategicOptionsfor BangladeshReadymadeGarmentIndustry I.Introduction 1.1 The end of M F A quotas signals intenseglobal competitionand enormouschallengesaheadfor Bangladesh's readymade garment (RMG) industry. Undeniably, it forces the sector to actively search for measures to improve its competitive position by undergoing necessary adjustments fast. At the same time, the policy and physical environment under which the industry operates has to improve, if it i s to retain or enhance its share in export markets. The cost o f inaction or policy mistakes could be heavy. Hence, the urgency to explore strategic options and put them into operation. 1.2 The RMG sector has become the lifeline of the Bangladesh economy - for jobs, income, and exports. Some 10million people, directly or indirectly, depend for their livelihood on this industry,which accounts for 40% o f industrial employment. It i s o f strategic importance to strengthen its competitiveness through appropriate early actions, and to resist pressures to retain policies that are not in the overall interest o f the nation. The sector has played a far greater role in Bangladesh's growth performance and employment generation than what might be inferred from this'sector's low share in GDP (barely 5%). This very dynamic sector has become the main source o f direct and indirect employment, foreign exchange earnings, which in turn helped finance a growing share o f imports o f vitally important capital goods and essential inputs. It has created significant additional positive externalities, including by being an important training ground for a growing number o fnew entrepreneurs. 1.3 There is urgency to become more competitive and diversified. There i s indeed an urgency to act fast in addressing the priority policy and institutional constraints to improving Bangladesh's overall competitiveness and that o f the RMG sector. All economically sensible options to further improve competitiveness o f the RMG sector need to be pursued. Bangladesh cannot afford to (and should not) let the RMG sector lose its international competitiveness. The fact that RMG exports make up over 75 percent o f the total export basket gives rise to certain vulnerabilities under the post-MFA global regime. Export concentration, in and o f itself, presents a diversification challenge. But with the phase out o f the MFA, and the consequent competitiveness pressures on the RMG sector, export diversification takes on new meaning for Bangladesh which must count on superior export performance inthe medium- to longer term for sustained high growth and reduction o f poverty, if the MDG targets on poverty and human development are to be attained. 1.4 Thepost-MFA external environment is beyond the sector's control. But domestic policies and the physical environment can be shaped, for positive results. Given Bangladesh's export concentration in RMG, a two-pronged approach is essential to meet the post-MFA challenge: (a) addressing and removing overall constraints to export competitiveness to unleash forces o f export diversification; (b) focusing on the key policy and institutional constraints relating specifically to the textile-RMG sector in order to seize opportunities for market expansion abroad and job creation at home. World Bank's (2005a) just completed Growth and Export Competitiveness Study was a response to the first challenge. The present study relates to the second and tries to identify the critical constraints to competitiveness o f RMGsector and recommends strategic policy options available with the public and private sectors to ensure competitiveness o f RMG exports in order to retain and augment Bangladesh's market share inthe global marketplace. 1.5 The World Bank (2005a) report examined export competitiveness challenges in a broader context, and was able to identify the generic as well as product specific constraints that undermine export 1 competitiveness - due to policy, institutional and infrastructural bottlenecks. The study revealed a number of key cross-cutting "Behind-the-borde?" constraints to export competitiveness. These include weaknesses in economic governance and transport-telecom-port infrastructure; high cost o f finance; cumbersome import regime and dysfunctional duty drawback system; product quality, consistency and standardization problems, poor labor slulls and low productivity. The present study takes on board the findings and critical recommendations o f that report before charting out a menu o f strategx options for the RMGsector under the post-MFA regime. 1.6 A globally competitive RMG sector is also key to poverty reduction. Bangladesh's Poverty Reduction Strategy Paper 2005 (PRSP) emphasizes acceleration o f pro-poor growth as the route to halving the proportion o f population living below the poverty line by 2015 - a key MDG target. Job- creating export expansion, based on the nation's comparative advantage, i s the fulcrum on which rests such a strategy for accelerating growth with poverty reduction. No other substantial manufacturing sector offers greater opportunity to increase employment for a givenvolume o f investment than the RMG sector. Hence, the justification for a special focus on this sector. 1.7 The layout of this report i s as follows: Section I1begins with a brief description o f post-MFA global changes in demand and supply sourcing for T&C products. Section IIIdiscusses the implications o f these changes for Bangladesh's global market share for RMG and their impact on domestic economic welfare resulting from likely changes injobs and output. The structure o f domestic industryand sources o f competitiveness are described in section IV, while section V lays down the strategic options for the industry in the face o f heightened global competition. Section VI concludes with a set o f recommendations for policy. 11. Post-MFAGlobalPerspective:Shape of Thingsto Come 2.1 Trade in textile and clothing (T&C) products has been managed under Multi-Fiber Arrangement (MFA) since 1974. The hallmark o f MFA was quota restrictions on imports into developed country markets o f T&C products from a number o f developing countries. The 1994 Agreement on Textile and Clothing (ATC), reached with the establishment o f the WTO, stipulated phasing out all quota restrictions under MFA by 31 December 2004 and the full integration o f T&C products into the multilateral trade disciplines. Accordingly, the quota restrictions were withdrawn by all countries by the agreed date, and T&C entered the uncharted waters of quota-free trade. In order to formulate the strategic options for Bangladesh, it i s essential that we have a good perspective o f the changed global scenario - markets and supply sourcing -- and their implications for Bangladesh RMG industry. 2.2 The world export of apparel doubled between 1993 and 2003, risingfrom about $90 billion to $180 billion. L o w and middle income countries shared in this expansion the most, their apparel exports rising from $53 billion to $123 billion, ending with a 70% share o f global exports. While countries like China and India were restrained, these countries (which included Bangladesh) benefited from the quota system as well as preferential access granted to many LDCs (e.g. by EU under GSP and Everything but A r m s , and, recently, by U S A under AGOA). ' The term `behind-the-border' refers to economic policies, institutional and physical structures within the country that affect economic activity and investment decisions, thus malung a distinction from those policies and institutional arrangements that directly affect external merchandise trade and other external transactions- such as trade policies andpolicies governing ports or customs procedures. 2 2.3 Preferential market access and ATC led to the allocation of important economic and human resourcesfor exportproduction, and thus to welfare gains (or losses), but it also led to increasing the dependency of developing countries such as Bangladesh on these trade preferences. Since EU and U S were the major markets protected by quotas, it i s expected that the main effects o f quota abolition will be felt inEU, U S and the countries exporting to these markets. 2.4 Table 1 presents selected countries with the highest dependency o f export o f apparel and textiles to U S and EU.It is apparent from the data presentedthat although the sector i s relatively a small share o f GDP, Bangladesh i s vulnerable to external shocks due to export dependence on one product category - apparel. Table 1: Export of apparelandtextilesto the US andEU Country Apparel Textiles Export of apparel Export o f apparel and as YOof as % of and textiles to U S textiles exports to exports to and EUas to U S and EUas EU-US EU-US share o f total share o f GDP (%) exports ("h) Cambodia 84 0 59 50 Maldives 50 0 86 43 Mauritius 54 0 59 33 Sri Lanka 50 1 35 18 Bangladesh 79 1 75 11 Pakistan 38 13 28 4 Nepal 29 13 34 4 India 16 7 9 1 China 6 2 3 1 Source: UNCOMTRADE and World Bank 2005. 2.5 The freeing up of trade in T&C products have important output, employment and trade implicationsfor both developed and less developed countries. When analyzing the end o f ATC3,it i s useful to consider three groups o f countries: 0 developed countries that used quotas to control the imports o f textiles and apparel (USA, European Unionand Canada); 0 countries whose exports o f textiles and apparel were constrained by quotas (China, India, Pakistan); and 0 a group o f developing countries that used the quota system as an opportunity to develop and promote their export of textiles and apparel (e.g. Bangladesh, Sri Lanka, Cambodia, Vietnam). 2.6 So what i s likely to happen after the conclusion o f ATC as o f end December 2004?A computable general equilibrium model, GTAF' (Global Trade Analysis Project; See Annex Ifor a brief description o f the model), was used (World Bank 2005) to determine the possible outcomes globally, as well as for Bangladesh. Because o f the high dependence Bangladesh has on RMG exports, there i s a strong presumption that quota abolition will have negative impact on its exports with serious output and employment consequences domestically. This section draws heavily from the 2005 World Bank policy note prepared for DECRGTrade by Vlad Malone. 3 2.7 The CGE techniques using the most recent data available4, provides estimates o f the impact on economic welfare of these countries considered under three scenarios: 0 quota abolition for apparel and textiles; 0 imposition o f safeguards by U S and EUon top o f quota abolition; and 0 imposition o f an export tax o f 2 percent on apparel and textiles by China on top o f quota abolition. A fourth scenario i s also explored: to see the potential impact on Bangladesh o f duty-free access o f its exports into the US market. 2.8 Though quota abolition was initially desired by all developing countries (first scenario), the impressive growth of apparel and textiles industry in China and the obvious quota constraints on Chinese exports (Martin, Manole and Van Der Mensbrugghe, 2004) led producers fiom developed and some developing countries to ask for safeguards measures from the U S and EU (scenario 2), which was then incorporated into China's WTO accession agreement.5 China proposed an alternative voluntary restraint measure: an export tax on apparel and textile sector (scenario 3). 2.9 The economic welfare outcomes for the three scenarios are summarized in Table 2 for selected countries and regions, including Bangladesh. Table 2: Economicwelfare resultsfrom GTAPsimulation (US% million) CountryIRegion Quota abolition Safeguards Export tax by China EU25 9358 -1581 -705 UnitedStates 6041 -1191 -598 China 3139 590 540 India -267 60 5 Bangladesh -370 93 57 Sri Lanka -142 24 10 Rest of SouthAsia -45 53 29 Other countrieshegions, -1670 87 -770 Total of a i regions. 16044 -1865 -1432 Source: World Bank 2005 2.10 The model results clearly reveal the benefits of quota abolition which, interms o f money-metric measures o f economic welfare, aggregates to over $16 billion, with principal beneficiaries being US, EU25 and China. As expected, Bangladesh loses out to the tune o f $370 million, but gains fiom any safeguards against China as well as Chinese export taxes. Much o f the welfare changes emerge from changes in allocative efficiency and in terms o f trade. However, it must be kept in mind that the comparative static results of the model are not necessarily contemporaneous but takes place over a period of three to four years, assuming Bangladesh is not addressing the competitiveness challenge through appropriate measures. The adverse impacts, ifany, would thus occur with a lag, leaving time and scope for corrective action for countries like Bangladesh who stand to lose out, ifnothingi s done. The last version of GTAP, GTAP 6.4 was used ( www.ptap.org ). See Annex Ifor a description o f the model and its limitations. These safeguardsstipulated restraints on China's textiles and apparel export up to 7.5% growth each year. 4 2.1 1 Following quota abolition, the reconfiguration o f export and import flows between regons and countries results in changes in employment o f unslulled labour, w h c h i s the principal category o f workforce engaged in textile and apparel industry across countries. The implications for employment changes are summarised in Table 3 below showing Bangladesh could suffer significant employment losses - 17 percent in apparel and 5 percent in textiles sector though India and perhaps Palustan achieve notable gains in South Asia. The Chinese workforce appear to be the major winners as demand for textile and apparel, including those of their linkage industries, rises sharply with opening up o f global markets. Table 3: ChangesinUnskilledLabour after QuotaAbolition CountryIRegion Changes inUnskilledLabor ("A) Raw Materials for Textiles Textiles Apparel EU25 2 -17 -19 China 24 26 59 Bangladesh 1 -5 -17 India 3 4 6 Sri Lanka 5 -3 -17 Rest o f South Asia 6 13 3 United States -1 -12 -19 Source: Vlad Manole (2005) Table 4: Export of RMG2001-05 (US$ million) Woven Knit RMG garments garments Total 2000-01 3,364.2 1,496.4 4,860.6 2001-02 3,124.6 1,459.2 4,583.8 2002-03 3,258.3 1,653.8 4,912.1 2003-04 3,538.1 2,148.0 5,686.1 2004-05 3,598.2 2,819.5 6,417.7 Year-on-year growth ("A) 2001-02 -7.1 -2.5 -5.7 2002-03 4.3 13.3 7.2 2003-04 8.6 29.9 15.8 2004-05 1.7 31.3 12.9 Source: Export promotion Bureau 2.12 Model predictions versus actual performance. MFA quotas were phased out at the end o f December, 2004. The GTAP model used predicted export shocks and economic welfare losses for Bangladesh, albeit with a lag. The immediate post-MFA period has not revealed this trend. RMG exports from Bangladesh in the post-MFA world did not decline; rather, data show that export growth during January-July 2005 was substantial to end the fiscal year 2004-05 with a growth o f 13% (Table 4), with knit garments registering phenomenal growth o f 31%. Indeed, performance in the hitherto quota- restricted market o f U S A revealedinherent strength o f the Bangladesh industryrelative to its competition (Table 5). DuringJan-Jul, 2005, Bangladesh was clearly a winner with China, India and Sri Lanka, while countries like Hong Kong, Taiwan and Malaysia lost out. Knitwear exports to U S jumped 70% during Jan-Jun 2005 and BKMEA reports that order books are Yet, this should give no reason for Part of the problem was the absence of quality data o n quota premia (export tax equivalents) in Bangladesh. Results cited here and their divergence from actual performance suggest that assumptions o f ETE were 5 complacency, as major challenges lie ahead: safeguards on Chma will be phased out by 2008; Indian textile and garment sector has undergone major technology uplift and policy reform to put India amongst leading contenders in garment exports; and inherent weaknesses in policy continue to plague the woven garment sub-sector, whose share in exports i s 56%, though declining. EU buyers are loolung at neighbouring Turkey, Tunisia and Morocco, as makers o f garments usingEUfabrics to meet the demands o f leanretailing and quick deliveries. Table 5: US Import of MFA Apparel, January -July 2005 ~ ~~ Jan-JulO4 Jan-JulO5 Growth Jan-JulO4 Jan-JulO5 Growth (million square meters) ("/I (US$million) (%> World 11,018 12,443 12.9 35,466 38,827 9.5 Bangladesh 512 613 19.8 1,056 1,284 21.5 Cambodia 324 352 8.6 777 888 14.2 CBI 2,27 1 2,369 4.3 5,559 5,626 1.2 China 1,536 3,414 122.2 4,673 9,08 1 94.3 HongKong 392 246 -37.2 1,934 1,444 -25.3 India 369 476 28.9 1,324 1,780 34.4 Indonesia 412 464 12.6 1,386 1,595 15.1 Malaysia 112 109 -2.9 390 353 -9.4 Pakistan 290 315 8.5 623 666 6.9 Philippines 284 275 -3.3 987 982 -0.5 SouthKorea 311 179 -42.5 968 616 -36.3 Sri Lanka 225 265 18.2 817 951 16.4 Sub-SaharanAfrica 245 220 -10.0 931 836 -10.2 Taiwan 304 202 -33.6 829 602 -27.4 Thailand 287 295 2.8 940 1,000 6.4 Vietnam 419 426 1.7 1,454 1,440 -0.9 Source: Office o fTextiles andApparel (www.otexa.ita.doc.gov) 2.13 The China Factor. There i s widespread concern that the end o f MFA would allow the more competitive countries, such as China, to develop the entire range o f textile industries from cotton to ready-made garments in order to take advantage o f the new trading regime. The surge in China's import o f textile and clothing related machinery appears to corroborate this concern. Between 2000 and 2003, China's import o f textile and clothing machinery increased from about $2 billionto more than $5 billi~n.~ This is being interpreted as suggestive o f China's desire to increase its market share very substantially. China i s known to produce high quality fabric. The modernization of the textile industry through large scale investment will allow it to further improve upon quality to meet the standards demanded by the global market. Availability o f domestic quality fabric will enable the RMG producers in China to meet tighter delivery schedules. With an increase in China's share o f the global market, there will be an incentive for migration o f textile capacity to China. Some experts are predicting that China may capture as much as 50 per cent o fthe global export market.8 underestimated, compared to, say, Pakistan's (13%). Assuming a higher ETE (e.g. 13%) and re-running the model yields results closer to actual performance - significant export volume growth inhtherto US quota market, modest declines in exports to EU, and lower economic welfare and employment loss. Post-MFA performanceinUS suggest that MFA quotas were more binding for Bangladeshthanpresumed. See OECD (2004). 8 Signs are emerging that China could indeedgreatly increase its market share. Chinese customs data for exports to USA show that exports of major apparel products (8 product groups) increasedby a massive 546 per cent during the first month after quota abolition over the correspondingmonthof the previous year. During January 2005 China shipped more cotton trousers and cotton knit shirts to USA than it had done inthe whole o f last year. Sucha large increase was made possibleby very large price cuts, 25 per cent on average and inexcess of 39 per cent for half of the 8 products (Press Release, National Council for Textile Organizations, March 7, 6 2.14 The anxiety about the dominance o f efficient Chinese producers in post-MFA world is reflected in the WTO protocol on the accession of China. It contains a transitional product-specific safeguard mechanism that allows WTO members to impose restrictions on imports from China if such imports are regarded disruptive o f the market (Article 16). There i s also a textile safeguardprovision valid till 2008. This provision has already been invoked by USA and EU to restrict imports o f certain categories o f apparel o f Chinese origm. The safeguard provision introduces certain amount o f uncertainty about the prospect o f unlimited market access o f Chinese export^.^ This should work out well for other countries including Bangladesh. Since there is substantial risk in relying on only Chinese T&C products, retailers and importers indeveloped countries would be inclined to diversify their sourcing to include several other competing countries. Bangladesh must ensure that it remains one of the major sources o f clothing products. However, it should not be lost sight o f that although the safeguard actions may provide temporary respite, they cannot be relied upon for the long term growth o f the industry; that will depend ultimately on competitive strength. If the Chinese price cuts reflect disappearance o f quota premia - which ranged from 25% to 45% -- then other countries will in the long term have to match the Chinese price cuts inorder to maintain their market shares. 111. TheDomesticContext:Challengesto Overcome 3.1 In this section, we examine how the Bangladesh RMG sector is positioned vis-$-vis the domestic economy and the coping mechanisms at hand for facing the onslaught o f post-MFA global competition. The sheer size o f RMG sector, which accounts for more than three quarters o f total exports, implies that Bangladesh will have to count on its superior performance in the medium to long term if the Millennium Development Goal (MDG) targets on poverty and human development are to be attained. Bangladeshi exporters including the RMG exporters, face enormous challenges to their competitive advantage. Given the problem o f export concentration in RMG, and the fact that the sector provides employment for nearly two million workers, mostly women, it appears to be vulnerable to the MFA phase out. There is genuine concern about the competitiveness o f Bangladesh RMG products relative to those from China, India and some other developing countries due to the sector's relatively l o w productivity, perception o f poor product quality and long leadtimes. 3.2 Many of the weaknesses of the RMG sector are due to cross-cutting infrastructure and governance related problems. The World Bank (2005a) study noted earlier emphasized the need to address these urgently in order to improve the efficiency o f the entire economy, and thereby enhance the export competitiveness o f the RMG as well as other export industries. The productivity o f the RMG sector i s also constrained by a lack o f adequately skilled manpower. Compliance with various social, ethical, health and environment related standards demanded by buyers i s putting extra strain on the competitive strength o f the sector. These issues have been analysed indetail by a study commissionedby the Ministry o f Commerce (Gherzi et aZ2002). The IFC-commissionedreport by Dr Martelli Associates (1999) looked at the entire supply chain o f the T&C industry, i.e. from spinningto garmenting, and came up with very similar conclusions. These studies, especially Gherzi et aZ, also made a detailedstudy of the forward linkage activities and pointed out the weaknesses o f the industry in this respect. There does not seem to be many irreconcilable differences regarding broad policy options to address the cross-cutting and some o f the sector specific issues such as compliance. However, there seem to be a number of yet unresolved issues related to backward linkage, lead time, central bonded warehouse and accessing GSP through regional cumulation. W e take a closer look at the economic arguments behind these issues in order to arrive at informedpolicy options. 2005, www.ncto.orq). US textile lobby is fiercely campaigningto have safeguard measures imposed onmost of theseproducts. China has imposed a small export tax on apparel ostensibly to moderate any surge inexports. 7 3.3 Thefact of the matter is that Bangladesh has come to rely heavily on its garment sector for jobs, income and foreign exchange. From its current position among the leading garments exporting nations of the world, Bangladesh cannot afford to be sidelined, as the economy-wide effect of this could be traumatic. It is, therefore, absolutely essential that every effort be made to not only retain its current competitive edge in the RMG sector but to enhance it by constantly improving productivity through, among other things, skill development, application o f cutting edge technology, courting foreign investment, and talung advantage of scale economies in production and exploiting any opportunities offered bybilateral, regional or multilateral trade concessions. 3.4 No doubt the unfolding global economy will impose harsh disciplines on the economy; but there will be also opportunitiesfor competitivesectors. To the extent Bangladesh garment exports have been sheltered fi-om competition in the past they will face the full force of competition from other low- cost suppliers such as China, India, Pakistan, and Vietnam. To the extent Bangladesh has been restricted by quotas and tariff barriers, or it has gained in competitive strength, it will benefit from greater opportunities than have hitherto been available inthe U S market. Box 1.Post-MFAActionProgramme (PMAP) The Ministry of Commerce has recently designed a Post-MFA Action Programme to support the RMG sector and mitigate possible negative shocks flowing from quota abolition. PMAP i s to be implemented infive years, at a cost of $40 million. The Programme has six components: Skill and Quality Development Programme: Training is given in several areas such as compliance, quality management and marketing inorder to improve the performance ofthe sector. Displaced workers Rehabilitation Programme: to assist and retrain those who might lose their jobs Support to Capacity Enhancement Programme: Ths has two sub-components - (a) to assist capacity enhancement o f smaller RMG producers by helping them form strategic partnerships, mergers, and productivity improvement programmes; (b) Technological Capacity Development Programme, to help SMEs inthe R M G and textile sector to access improved technology to enhance their competitiveness. Support for PTS to improve quality andreduce costs. Assistance for the handloom sector: to make them more competitive by setting up separate design and development centres for bothhandloom and PTS. Assist forward linkage industries to enable themto provide better service to the RMGsector. Funding from development partners has been sought to finance the PMAP. To start off, an allocation o f TK 200 million was made by GOB in the 2004-05 budget for re-training and employment o f potential retrenched garment workers, inaccordance with component 2 o f the PMAP. Although it is a timely attempt to address the post-MFA challenge, it will be seen that the PMAP i s rather limited in its scope and budget. MOC seemed to have designed the PMAP without inter-ministerial coordination mechanism builtinto it. It does not also address the sector wide fhdamental competitiveness issues such as lead time, logistics barriers, import regime, availability o f inputs at world prices and preferential access that are of critical importance for the sustainability o f the sector. Without addressing these basic constraints, the support programthat is envisaged inthe PMAP is unlikely to bear hit. By ignoring these issues the government is passing up an opportunity tc strengthen the external positionofthe sector. IV. Implicationsof Quota Abolition 4.1 Thephasing out of M F A on 31 December 2004 marked the abolition of quota restrictions on exports of textile and apparels from some developing country WTO members by the developed countries. This will undoubtedly have important effects on both the exporting and importing countries; however, the effects are likely to be more far reaching and widespread in countries that have excessive reliance on textile and apparels for export earnings or industrial employment. Since, Bangladesh depends excessively on RMG for export earnings (over 75 per cent) and about half o f its industrial employment originates inRMG, it i s potentially very vulnerable to any adverse outcome due to the possible changes in the current market structure. To ensure that the FWG sector can withstand the shock o f quota abolition, both the industryand the government will have to respond with appropriate policy reforms and measures. It is quite conceivable that with such response, the challenges of MFA phase-out can be turned into an opportunity that would lead to greater expansion of the sector as well as other linkage industries. 4.2 The effect of quota abolition will, in thefinal analysis, depend on the competitive strength of the sector. If it has already attained international competitiveness, or can quickly attain it, then the abolition o f the quota regime will turn out to be a welcome event that will promote further expansion o f the industry. On the other hand, if it i s lagging in terms o f competitiveness it will find its market share eroded due to encroachment on its market share by more efficient producers from other countries. 4.3 Most analysts regard Bangladesh as 'vulnerable'. A reason for this prediction i s the fear that large efficient producers o f textiles and clothing such as China, which were quota restricted in developed country markets, in particular U S A and EU, will severely undercut prices that will drive out the less efficient producers from these markets. The large quota premium on Chinese exports would permit a large undercutting o f prices following quota abolition. Some predict that China could secure as much as 50 per cent o f the market share. Market developments with respect to apparel items that were freed from quota restrictions in EU and U S A in 2002 as part o f ATC are usually cited as early indications o f what may happen in 2005 and beyond. Prices o f imports o f quota-decontrolled apparel items from China fell by 44 per cent in the US and by 42 per cent in EU (see M o f C 2004). US imports from China in the decontrolled categories rose by 290 per cent while EU's rose by 162 per cent. U S imports from rest o f the world declined by 14per cent during the same period." InJanuary 2005, Chinese apparel exports to U S A surged by over 500% with prices decliningby over 25 -40%. Such export performance, ifit were to continue, bodes illfor vulnerable economies such as Bangladesh. 4.4 The Chinese price reduction reflects the elimination of quota premium. Since the quota premium i s an estimate o f the excess o f export price over cost, it gives a rough indication o f the competitive strength of the producers of the exporting country. If quota restrictions are removed, producers could reduce price by this amount and still maintain output at the quota level." Any lesser reduction will witness an increase in sales inthe hitherto quota-restricted market. Hence, much attention has been paid to quota prices inthe empirical studies on the impact o f MFA phase out on the RMG sector o f various countries. lo The surge in Chinese exports to USA in some quota categories during the first month after the expiry of the MFAlends some credenceto the concern. Also see footnote 5. l1 Ifthe exporting country is also exporting to unrestrictedmarkets, it will not be able to reduce export price by the full amount of the quota price without reducing sales in the unrestricted market unless it operates under constant cost conditions. 9 4.5 A brief assessmentof Bangladesh'sRMG exports to its main quota market (USA) is in order. Inthe USA, Bangladesh RMG exports faced quotas like everybody else and got no GSP benefits. Moreover, these exports are subject to high tariffs12. In value terms, about 46 percent o f Bangladesh's exports o f clothing products are subject to an ad-valorem duty o f 15.1-20 percent. Another 13 percent faces tariffs higher than 25 percent. Hightariffs considerably reduce Bangladesh's competitive strength in the US market since as many as 72 Afncan and Caribbean countries get zero-tariff access in the US market under A h c a n Growth and Opportunity Act (AGOA) 2000 while Mexico gets the same under North American Free Trade Area (NAFTA). In order to formulate strategic options for Bangladesh's RMG exports, it is useful to make an assessment o f the changing US market where it faces intense competition. The United States International Trade Commission's 2004 Report on Textiles and Apparel: Assessment of the Competitiveness of Certain Foreign Suppliers to the US Market presents the likely effects o f MFA quota removal in U S market on Bangladesh vis-a-vis its three major competitors viz. China, India and Pakistan. The report also lists what might be the key contributing factors (Table 6). As described earlier (Table 5), during the Jan-July, 2005 period, Bangladesh exports have faired better than analysts' predictions so far. If Bangladesh were to get zero-tariff access into the U S market - if the proposed U S TRADE Act were to be enacted - estimates suggest that its exports could double over the next three years. Table 6: Summaw of anticipatedeffectsof auota eliminationinUS market in 2005 ahd key competitivefa ors; by selected countries Country Likely effects of MFA quota removal Likely to be the supplier of choice for most large Labour Per-unit labour costs very low due to low wages and high - Contributingfactors US. apparel companies and retailers; uncertainty productivity. regarding textile-specific safeguards may temper Inputs Producesfabrics, trim, packaging, andmost other components - export growth. Over the long term, competitiveness usedto make apparelandmade-uptextile articles. - China may diminish as strong economic growth leads to Products Considered by industry among the best in making most greater domestic demand for textilesand apparel, and garments and made-up textile articles at any quality or price level. for the labour andcapitalto makethese goods. World's largest producer and exporter of textiles and apparel, notwithstandingtight quotasinmajor world import markets. Showed tremendous growth in export of goods for which it becameeligible for quota-freeentryin2002. Likely to remaina competitivesupplierto the United Labour Huge, relatively inexpensive, skilled workforce; has design - States when quotas are removed in2005. Considered expertise. by manyUS.firms the primary alternativeto China. Inputs Amongthe world's largestproducers of yams andfabrics. - Products Wide range of apparel; considered a competitivesource for - India Over the long term, competitivenessmay diminish as hometextiles (e.g., bedlinens andtowels). strong economic growth leads to greater domestic Business climate - Personal safety, security of shipments between demand for textiles and apparel, and for the labour factories and ports and bureaucratic red tape and infrastructure are andcapitalto makethese goods. issues, with many US. firms using agents in lieu of dealing directly with producers. Likely to continue as a supplier to the U.S. market. Labour Large,relatively inexpensivelabour supply. ~ Considered by many US. firms as a competitive Inputs --Access to local supplies ofraw cotton. Pakistan alternativeto China, particularlyfor men'sapparel. Businessclimate The government i s taking steps to ensure the global - competitiveness of the textile and apparel sector; personal safety and May continue to be a global supplier of cotton yams security of shipmentsbetweenfactoriesandports are issues. andfabrics. The status of Bangladesh as an overall supplier to Labour - Very low wage rates; productivity improving,but lags China; US. market is uncertain. Considered by some U.S. governmentisworking to improvelabourstandards. firms to be competitivealternativeto Chinafor mass- Inputs - Relies heavily on imports for woven fabric requirements; Bangladesh produced, low-endapparel. becomingincreasinglyself-sufficient inknit fabrics. Special arrangements - Duty-free access to major world import markets, including the EU, Canada, andNorway. Products Mass-produced basic garments, including knit cotton tops - andwoven cottonpants. Source: Quoted from USITC (2004) I t is on record that in2003 Bangladesh paidmore tariff revenues on its $2billion o f exports than France didon its exports of $26 billionto the US. 10 4.6 Thefinal impact of quota abolition on individual countries depends on an array of factors in a complex way. When all these factors play out, a new equilibrium price will be established inthe market. Whether a country's exports increase or decline will depend on its cost relative to this new price (see Appendix B for a technical discussion). Hence, the key to success in the post MFA regime i s to reduce cost relative to other countries, i.e. to be competitive. Reducing cost relative to other countries is particularly important in view o f the fact that many o f these countries are reorganising their textile and RMG industry through appropriate policies, new investment, technological improvement, forward and where possible backward linkages and conducive business environment in order to exploit the opportunities that will present themselves with the full integration o f the industry into multilateral trade agreements. Any country that lags behind in the global competition race will be penalised by a reduction o f its market share. Bangladesh must ensure that its RMG industryi s up among the winners in this race, and this can be achieved only by reducing costs and satisfying the demand o f the customers in terms of lead time, design, quality and standards. 4.7 Bangladesh has achieved a global reputation as a reliable supplier of low-value basic items of apparel. Despite binding quota restrictions in the U S market in the past, Bangladesh has remained a supplier o fthe cheaper products as it apparently has a comparative advantage inthe production o f the less expensive clothing. Annex Table 1provides unit values o f selected items of some developing countries to U S A before quota phase out. It clearly shows that Bangladeshi exports are among the least expensive; only Pakistan offers cheaper prices across a range o f products. Of special interest are the prices o f export products o f China and India, which are regarded by many as the principal threat to Bangladesh and other apparel exporting countries. Contrary to the general perception, Chinese products are seen to be far more expensive than the products of Bangladesh and other countries listedinthe table. Indeedthe unit values of the Chinese products are on average thnce as expensive as the unit values o f the Bangladeshi exports. 4.8 Note that unit values derived by dividing total FOB value o f exports by the volume may not be a good indicator o f the cost conditions because o f the distortions introduced by quota restrictions o f the past. These realised market prices would have then included quota premia if the quotas were binding for any country. Hence, the quota effects needto be netted out to find the producer prices that would exist in the absence o f quota. Annex B technical discussions show that quota premia acted as an export tax which exporters hadto recoup inorder to stay engaged inexport business. 4.9 Annex Table 2 provides estimates o f quota prices o f different categories of apparel in selected counties. As will be seen, quota prices varied quite markedly. They were the highest in China which had the highest prices for almost all categories. India has on average the second highest quota prices. Quota prices in Bangladesh were on average about the same as in the other countries in the table.13 Assuming that quota prices given in Annex Table 2 are comparable, it i s possible to get an estimate o f the net producer price o f an item by subtracting the quota price from the market price. This is done in Annex Table 3 for Bangladesh and China. It will be seen that adjusting market price for the quota premium only marginally reduces the proportion by which average Chinese price exceeds average Bangladeshi price. The ratio o f Chinese and Bangladeshi average net price falls to 2.7 reflecting a higher quota margin for the Chinese products. The lower this ratio, the more competitive the Chinese exports relative to Bangladeshi exports. 4.10 Since quotas applied to volumesrather than values, Chinese suppliers must havefound it more profitable to concentrate on high value items. That explains the generally higher price of Chinese exports to the U S market compared to exports from Bangladesh. Indian and Sri Lankan product prices are l3 According to the estimates of Elbehri (2004) the export tax equivalent of quotas o f Bangladesh i s 70 per cent higher than that of China. However, other estimates reported inRazzaque (2005) are similar to that reported in t h ~ paper. s 11 also considerably higher in most cases. It would appear that Bangladesh serves a cheaper segment o f the U S market than these countries. Hence, Bangladesh may not be in direct competition with them. However, it i s possible that these countries could foray into the lower end o f the market which earlier they had not because o f the quota restrictions. To what extent this climb down from high end to the low end o f the market would be successful is difficult to predict. Those who produce high quality, high value products efficiently may not necessarily be competitive inthe market for cheap products. V. IndustryStructureand Sources of CompetitiveAdvantage 5.1 Industry structure. Before exploring strategc options for the industry, it makes sense to take a quick look at the industry structure and the scale economies it presents. Bangladesh Garment Manufacturers and Exporters Association (BGMEA) Members ' Directory 2004-2005 lists 4300member firms with the following breakup: 700 knitting factories, 525 sweater factories, and the remaining 2275 are woven garment units.However, a BGMEA internal survey found that some 1300 o f these firms were closed or inactive. In addition, Bangladesh Knitwear Manufacturers and Exporters Association (BWEA), boasts a membership o f another 860 knitwear firms, o f which some 300 are believed to have dual membership with BGMEA, leaving 560 units with exclusive BKMEA membership. Thus the structure o fthe garment industrywith active firms may be summarized as follows: Table 7: Structure ofRMGindustry Garment category Numberoffm Share (%) Empl~yment'~ (adjusted) Woven 1673 47 836,500 Knitwear 1495 42 747,500 Sweaters 392 11 337,120 Total 3560 100 1,921,120 4.2 Thus, just about half of theflrms are into woven garments, while knitwear and sweaters together make up the other half of the industry. The latter group has shown tremendous buoyancy o f late, with relentless growth particularly in the EUmarket, notwithstanding the end o f the quotas. Capital investments per unit of labor employed i s relatively l o w for the knitting firms. Woven firms are typically larger -- about 13% of them also engage in the knitting industry. These are usually the larger and more productive woven firms. 4.3 Although'BGMEA and BKMEA data on member firms would yield total employment o f about 2.8 million workers in 2004, adjusting for inactive firms and using our latest survey o f 350 RMG firms, yields a total employment o f 1.9 million productionworkers, or 2.0 million, ifone includes all employees. At least another million are engaged in accessories and linkage industries (e.g. packaging, courier services). 1% o f garment firms operate in EPZ, 63% o f which have foreign ownership. These firms use more capital and technology and have higher productivity than those outside, as i s reportedlater. Data from Export Promotion Bureau on RMG exports reveal the degree o f export concentration at the firm level. As it turns out, o f the 2387 exporting RMG firms in2004, the top 500 f m s exported 74% of total garment exports; the top 650, up to 81% (Table 5). Which means the remaining 1737 firms in EPB's list o f exporters managed only 19%between themselves. That leaves some 1200 firms (out o f 3560) that did not export at all or perhaps sub-contracted. Among these firms, some could be facing closure or l4 Because o f the problem of inactive fm,these employment estimates are based on the average number of productionworkers found in the 350 RMG fm level survey which followed stratified sampling procedures (see Table 8). 12 operating intermittently. Among exporting firms, the bottom 200 firms which export less than $50,000 eachcannot also be open for business all year. Cumulative export share (YO) Top 100firms 40.5 Top 200 firms 53.0 Top 300 firms 61.9 Top 500 firms 74.3 Top 650 firms 81.0 Box2. IndustryStructure andRMGExports A Ministry o f Commerce study (2001) found that the RMG sector comprises the following four groups o f 'm:Thefirstgroupofabout 15largecompanies own220manufacturingunits, eachwithacapacityof10,000 lozens or more per month. These are well established firms that source their own fabric and have direct marketing ink overseas. Another 550 manufacturing units have a production capacity o f 5,000 to 10,000 dozens per month who work mostly (60 per cent) on a CM basis for importers or buying agents. There are an estimated 1993 units :ach with a capacity of less than 5,000 dozens that work mostly on a sub-contracting basis. Finally there are about 200 units that are sick andnot currently operating. Duringthe early years the RMGindustrycomprised almost entirely o fwoven apparel manufacture. Evenas late as 1993-94, only 17 per cent o f total RMG export was accounted for by knit apparel. Since then the knit sector y e w very rapidly. Duringthe ten year period 1993-94 to 2003-04, the annual growth rate was in excess o f 23 per :ent. Duringthe same period the woven apparel export increased by a more modest rate o f 10.6 per cent per annum. As a result o f the differential growth rates the share o f knit apparel intotal garment export increased to 37.8 per cent while that o f woven apparel decreased from 83 per cent to 62.2 per cent. Duringthe first 8 months o f the fiscal year 2004-05 the share o f knit increased further to 43 per cent with an equivalent decrease in the share o f woven. Assuming this trend persists, in two to three years, the value o f export o f knit apparel will exceed that o f woven apparel suggesting relatively greater international competitiveness o f the former relative to the latter inthe world market. There is also high spatial concentration inthe export o f the apparel products. More than 79 per cent o f the knitwear exports went to the EUcountries in2004-05 while the U S A absorbed 14per cent. Canada took inanother 4 per cent. Thus these three markets accounted for nearly 98 per cent o f the total kmtwear exports o f Bangladesh There was similar concentration inthe case o f woven exports. These three markets imported 98 per cent o f the tota' woven exports o f Bangladesh. EUimported47 per cent, U S A 45 per cent and Canada 5 per cent. Woven exports are thus more evenly distributed among EUandUSA unlike the knit exports. There are other potential markets in the horizon. Japan, which is the largest importer o f RMG, afte: U S and EU, is yet to be explored, but holds enormous potential. Bangladesh needs to penetrate this market, but face! stiff competition from China. Vigorous marketing andpromotion drives are needed to enter the Japanese, East Asiax and other growing markets o f middle income countries, not to mention Australia, N e w Zealand, and Norway, all o which have given zero-tariff access to Bangladesh exports. 5.5 Employment in exportingJirms reveal certain interesting characteristics. EPB firm-level data reveals that employment in the 2387 exporting firms adds up to over 1.2 million (or 60% of total RMG employment). Which means that some 0.8 million o f the RMGworkforce are employed infirms that sub- contract (no direct exports) or belong to the group of firms that are vulnerable. Further, the top 650 firms that export 81% of total RMG are found to employ about 411,000 workers (21% o f RMG employment), while the remaining 1737 firms that export only 19% o f RMG employ around 790,000 workers (40% o f total). Clearly, this means that the workforce inthe latter group are terribly unproductive (big concern for the sector) or are not fully employed. 13 5.6 Sources of competitive advantage of RMG. Having looked at the RMG industry structure, we examine the sources of competitive advantage of the sector which derives from three channels: (a) factor costs, (b) domestic policy measures and (c) external trade environment. Each of these contributed significantly to the stellar performance ofthe sector during the last quarter of a century. Table 9: Labour and power cost in selected countries Labour cost per hour US$ Power cost U S cent/kwhr China 0.69 7.10 India 0.58 8.33 Indonesia 0.32 3.65 Mexico 2.20 n.a. Pakistan 0.37 6.00 Sri Lanka 0.46 7.78 Vietnam 0.27 n.a. Bangladesh 0.30 6.75 Source: Gherzi et a1(2002) Table 10: Estimated cost of investment Spinning Weaving unit Woven fabric Knittingand Woven Knitgarment (ring) unit processing lautfabric garment unit unit unit processing unit output Yam Grey fabric Finished Finishedlait Garments** Garments* woven fabric fabric Annual capacity*** 5.2 15.3 24.0 13.2 250,500 500,000 millionk g million million million metres dozens dozens metres metres Investmentcost (mil 756.6 555.6 742.6 267.0 161.4 103.3 taka) o f which foreign exchange 582.5 396.2 548.7 194.3 73.8 38.0 Employment 625 335 325 160 1250 745 Capital cost per 1.211 1.659 2.285 1.669 0.129 0.139 worker (million taka) *Ladies, Polo and T- shirts, and children's dress set; **Men, women and boys shirts and trousers; ***1 kg yam=6 metres of fabric; 1dozen woven garment-18 metres o f fabric. Source: Gherzi Textile Organisation(2002a). Executive Summaiy of the Initial Report on Post-MFA Development Strategy and TechnicalAssistancefor the RMG Sector, Report submittedto Ministryof Commerce, Government ofBangladesh 5.7 Factor costs: Ready-made garment manufacturing is a highly labour-intensive operation that requires relatively modest capital investment. A densely populated and capital-scarce country such as Bangladesh i s thus ideal for the expansion of RMG production. The skills required for making fast- moving and inexpensive garments are not too demanding and can be mastered by the nimble women workers of the country without great difficulty. A very large pool o f poor women engaged in low productivity home work provides a ready supply o f unskilled workers who could be trained into reasonably skilled hands for garmenting. The pressure o f the unemployed labour force and the very low standard o f living allow the garment wage in Bangladesh to be kept at a level that i s one o f the lowest among competing countries (see Table 6). The modest capital requirement enables a large number o f 14 entrepreneurs to enter the industry (Table 7). Successful entrepreneurs find it relatively easy to expand their operations by reinvesting profits, and thus avoid costly bank loans." This reduces capital costs of RMG operations.I6 These are the principal strengths of the industry that can be maintained as long as there i s an excess supply of workers. Box 3. Comparative advantage: Basicsversus highvalue added apparel Most o f the RMG fm inBangladesh (more than 2000 i.e. two-thirds) offer only stitching capacities o n what is known in the industry parlance as cutting and making (CM) or cutting, making and trimming (CMT) basis (Industry leaders claim that share of CMT i s m u c h lower with increased use of FOB system). This mode o f operation requires the least amount o f skill and information about the international market, and is thus well-suited to the (limited) capacity o f this group o f RMGmanufacturers. Hqwever, being the least skilled, they are vulnerable to the vicissitudes o f the market, and may find it difficult to withstand adverse shocks. There is a natural progression inthe production o f apparels from low to highvalue items. Initially entrepreneurs start with basic items that require little sophistication on the part o f either the labour force or the management. As the management and workers gain inexperience o f the apparel market, their efficiency improves. Some o f them develop direct marketing strategies, and some beginto manufacture higher value-added items. The process has already commenced inBangladesh. The more successful entrepreneurs have established direct contacts with their ultimate buyers (retailers) and have developed sophisticated marketing strategies. Some o f them have already made a successful transition to high value apparel product such as dresses and suits. However, this should not leadto the suggestionthat the RMGmanufacturers should abandon the basic itemmarket and shift to the hgh value fashion items o f clothing. The former constitute by far the largest section o f the market, and abandoning this market wouldbe a bigmistake. Bangladesh RMG manufacturers have successfully contested the market for standard items for a fair amount o f time and its share has increased over time. This i s a clear indication that the country has a comparative advantage inthe production o f cheaper standard items (e.g. see USITC (2004) assessment). l l u s is still the basic strength o f the domestic RMG manufacturers and most o f the export earnings still derive from this source. As long as this strength i s obtaining, the manufacturers should not abandon the market, indeed they should exploit it to the fullest possible extent. Those who find an opportunity to earn higher profits by movingup the value ladder should o f course do so. This decision must be based on the profit calculus o f each entrepreneur, and not on such concepts as "higher value addition". High value added products are not necessarily high total profit products and there are no special advantages inproducing them. Countries concentrating on the market for highvalue apparels happen to be the most technologically advanced, skilled and efficient producers o f apparel items. T o secure a niche inthis market, Bangladeshi RMG producers will have to compete against the highly sophisticatedapparel manufacturers o fthe developed countries and such developing countries as China, Sri Lanka and Turkey. T o be successful inthe high value end o f the market the domestic RMG producers will need long preparation with gradual build up o f skill, technology and capacity. Alternatively, t h ~ scan be speeded up through intake o f FDI (as in China, Sri Lanka and Cambodia). It is doubtful if many o f them will attain that capacity in the near future. Much investment inhuman capital and technology will be neededto make the transition. Foreign direct investment may act as a catalyst inthis respect. 5.8 The low wage of workers in Bangladesh is unfortunately accompaniedby their lowproductivity which erodes part of the benefits of cheap labour. Despite the many years, adequate productivity improvement has been lacking in the sector (discussed in Section VII). M o r e efforts and investment are needed in order to raise the productivity of its labour force to place themselves in a superior competitive Debt exposure o f RMG industry is l o w according to the Bangladesh Garments Manufacturers and Exporters Association (BGMEA) leadershp. l6 Capital cost of borrowed funds is as hlgh as 10.5-14.5 per cent. There i s a large spread (about 5 per cent) between borrowing andlending rates of banks. 15 position.I7 Government intervention might be warranted to develop a high quality workforce through the creation o f education and training facilities. This i s one area where there i s wide scope for private-public partnership and co-operation. Government support i s warranted by the public good nature o f investment inthis area. 5.9 Domestic policies: Given the heavily protected trade regime of past decades, the rapid growth of RMG would not have been possible without some compensating support mechanisms. For one, the textile sector was heavily protected; imports o f many textile products were prohibited while others carried the highest tariff rate. Amidst this protective environment, the RMG industry needed to source inputs, such as yam and fabric, at world prices in order to compete in the world market. An early support mechanism was the duty drawback system; but this involved upfront payment o f duties on imported inputs that tied up substantial funds o f manufacturers besides involving cumbersome procedures for reimbursement. In order to relieve the garment manufacturers of these difficulties, the government introduced the system o f bonded warehouse in the early 1980s. The bonded warehouse facility was a significant business-friendly policy support instrument that eliminated the duty-payment requirement and also substantially reduced bureaucratic hassles and delays." The entrepreneurs in RMG now faced essentially the same input cost as their competitors in other countries. Essentially the system created a virtual free trade enclave for imported inputsfor the export-oriented RMGproducers inorder to level the playing field that was earlier skewed against them. The removal o f the protection-related costs enhanced the competitiveness of the sector and contributed to the rapid expansion o f its share of the international market for garments. 5.10 Another policy support mechanism was the introduction of the back-to-back letter of credit (LC) system. The garment manufacturers were allowed to import their inputs (fabric and accessories) under this LC system. When a manufacturer received an LC from an overseas client for the supply o f certain amount o f garments, this master LC could be used to open another LC in a local bank for the import o f the necessary inputs.The cost o fthe importeditems along with interest and other charges would be deducted by the local bank from the proceeds o f the sales o f the final output. Hence, the manufacturer was spared the financial involvement in the purchase o f the imported inputs, which often amounted to more than 65 per cent o f the total cost o f the garments. The financial outlay requirement for garment manufacturing was accordingly greatly reduced to essentially wage, energy, transport and overhead costs. This was especially useful for a large number o f small and medium enterprises (SME) that had neither large finance capital nor the collateral to secure large bank loans. 5.11 The importance of appropriate domestic policies inenhancing the competitiveness o f an industry i s underscored by the success of the aforementioned poli~ies.'~However, there are other policies, such as hightariff andnon-tariffbarriers to import, that actually reduce the efficiency andcompetitive strength of the domestic industries, particularly export industries that have to sell their products at international prices. Furthermore such protective barriers also introduce an anti-export bias in the economy by providing relatively greater incentives for import substitution. They also make the import regime for " Recently BGMEA has set up a modest fashion institute. This is absolutely insufficient for the needs o f the industry. It is also open to abuse by the RMG entrepreneurs. It has been alleged that there is widespread leakage from the individual bondedwarehouses. l9 Some piecemeal measures were taken to improve port procedures that also add to costs. Significant improvement incustom procedures reduced signature requirement for export consignments from 17 to only 5 thereby reducing the time required to complete formalities. All custom procedures are now completed within two hours. The cost o f freight forwarding has been reduced to only Tk 1250 per consignment from Tk15,OOO-20,000. A public-private committee has been formed to monitor the functioning o f the reforms. 16 exports rather cumbersome2'. It i s worth mentioning that the Government has taken some steps in the FY06 budget to ease the situation by reducing duties on dyes and chemicals, eliminating duties on machinery spares, in addition to continuing with the 5% cash subsidy on exports using local inputs. All this adds to the VAT exemption givento utilitycharges incurredby 100% exporters. 5.12 External trading environment: There is no denying that the rapid rise of the RMG sector in Bangladesh had much to do with the institution in 1974 of the Multi Fibre Arrangement (MFA)by the developed countries to protect their own textile and apparel industry against competition from the more efficient producers in developing countries, particularly in Asia. Quota restrictions were applied on developing countries that exported significant quantities o f textile and apparel products to the developed countries. 5.13 Quota restrictionson the efficientproducers of garments meant that the importers and retailers of garments in developed countries were forced to source garments from a large number.of less efficient countries. Producers in quota-restricted countries also started shifting their production facilities to countries that were still not so restricted. A large number o f developing countries embarked on garment manufacturing to feed the large clothing market o fthe developed countries. 5.14 Thefirst major export oriented RMG industry in Bangladesh was established soon after the coming into force of MFA in collaboration with Daewoo, a large garment manufacturer o f a quota- restricted country, South Korea. There was no quota restriction on the apparel exports o f Bangladesh to U S A until 1985 as its exports were not substantial. But quotas were imposed inthat year after triple digit growth rate during the previous five years, with exports rising to $150 million (1 per cent of U S market).21 Growthrate o f exports to U S A fell o f f (and quotas became binding),buttotal exports increased robustly such that by 2000-01, Bangladeshhad captured 3.3 per cent o f the U S market.22 5.15 EU offered even better tradepreference to the least developed countries (LDCs) that included Bangladesh. It granted quota-free (as well as duty free) access to L D C products. Although exports to EU were much less than that to U S A throughout the 1980s, the picture changed in the 1990s. By 2002-03, export o f garments to EUwas 43 per cent greater than that to the USA. The faster growth o f exports to EU was no doubt helped by the absence o f quotas that restricted exports of competing countries. The Everything But Arms GSP scheme introduced in2001 allowed virtually all L D C export products to enter the EUmarket at zero tariffs when other countries had to pay full or concessional tariffs.23This provided a considerable competitive edge to LDC garment exports that qualified for duty-free treatment inthe EU market since the MFN(most favoured nation) tariff rates on RMGproducts were generally quite high. 2o Imagine how simple life would be for RMG exporters if there were n o tariffs or import restrictions on textiles. There would be no need for "bonded" warehouses, no complaints o f leakages, no customs corruption o n this count. But the question is, can the primary textile industry survive such a "free trade intextiles" regime? The general consensus i s that, inits current state o f competitiveness, it probably cannot. 21 Islam (2001), p.51. 22 The market share fell o f f fractionally since then. 23 As the title o f the scheme implies Armament was excluded from the scheme. A few agricultural products such as banana, sugar and rice were given limited duty-free access untillate inthe decade after which the restrictions would be entirely removed. 17 VI. Competitiveness and Strategic Sourcing Decisions 6.1 M F A quotas gave market access and shelterfrom competition to many developing countries, including Bangladesh. Quota abolition removes that passport to export markets. Instead, market share can only be gained or retained via international competitiveness o f which price i s only one -albeit critical -element. DuringtheMFAregime,buyingcompaniestypicallylookedformorethanthelowestpriceof garments in malung their sourcing decisions. Quality consistency, speed to market, reliability and flexibility, and service were important attributes amongst suppliers that won them business. 6.2 I n the emerging competitive apparel market of the post-MFA world, importers and retailers will rationalise their sourcing practices in order to better position themselves in a fast changing market. A recent study commissioned by the Commonwealth Secretariat (Lezama et a1 2004) on sourcing practices of international buyers finds that major buyers o f apparel are increasingly adopting sourcing strategies that seek to minimise their total costs throughout the supply chain rather than minimizing the cost o f procurement from the RMG factories who are the lowest down inthe supply chain. This relentless drive for total cost minimization by overseas buyers poses certain challenges for the RMG sector o f Bangladesh. These relate to (1) price competitiveness, (ii)quality, (iii)social and environmental compliance and (iv) lead time. 6.3 Thefirst strategic shift in global sourcingpost-MFA is likely to occur with a shrinkage of the ``span of control." Buyinghouses would like to limit the number o f countries and factories from which they source their apparels. But what i s also evident i s that they will not be putting all their eggs in the Chinese basket. In making country choices post-MFA, macroeconomic and political stability, infrastructure, compliance standards, and good partnerships, will be critical. Globally, therefore, a lot many countries which held significant share o f apparel markets will face shrinkage or total elimination. According to various assessments, vulnerable countries include Nepal, Mauritius, Maldives, Sri Lanka, Thailand, Philippines, Poland. Most analysis [World Bank (2005b); USITC (2004); Business for Social Responsibility (2004)l predict challenging times for Bangladesh though all indications are that it will remain a major garment exporting country. 6.4 Beyond price, quality consistency, and turnaround time, a large body of factors govern competitiveness to determine sourcing decisions of buyers. In 2004, the World Bank participated in a MFA Forum for managing the transition to a Responsible Global Textiles and Garment Industry, along with other multilateral organizations (UNDP), labor groups (AFL-CIO), NGOs (Oxfam), and private sector (Business for Social Responsibility). According to research carried out under the auspices o f the worlung group, companies consistently stated that several significant factors influenced their sourcing decisions. Much the same views were gathered from the survey of apparel industries by the Commonwealth Se~retaria?~. 6.5 These sourcing criteria are likely to remain important in the post-MFA regime. A comprehensive list, withno order o fpriority, would include: 0 Social and environmental compliance 0 Strong relationships between buyers and suppliers 0 Preferential trade agreements 0 Pre-production assistance 0 Transportation: access, costs 0 Political environment: safety, corruption 24 Lezama, M., B. Webber, and C. Dagher, Sourcing Practices in the Apparel Industry. Commonwealth Secretariat, London, 2004. 18 0 TradeRegulationshawiers Raw material/fabric access, supply 0 Vertical integration 0 Service and Flexibility 0 Communicationskills 6.6 How does Bangladeshfair in an assessment of its attractivenessas a sourcing destination?On the positive side, discussion with buyers reveal that they are not just ready to abandon suppliers with whom they built long-term relationships over the years and even invested in skill development, management practices and workplace conditions. Yet Bangladesh presents a mixed bag o f strengths in some o f the above factors while weaknesses in others. The message for the industryis that low costs and industry efforts will keep it alive but they have to go beyond price and cost competitiveness in order to develop other sourcing criteria. Due to the perception o f weak social and environmental compliance in Bangladesh, buyers are now puttingextra emphasis on meeting L O codes regarding working conditions and labor practices as well as adhering to international standards o f waste management. Infrastructure weaknesses, lack o f raw materials, and antiquated banking system are other problems cited by companies as industry deficiencies that need to be overcome in the days ahead. This can only be done through concertedpublic-private partnership going forward. VII. StrategicOptionsfor BangladeshRMG 7.1 Thepresentstudy takes on board thefindings and recommendations of the Bank's Growth and Export CompetitivenessStudy (World Bank 2005a) before charting a course o f strategic options for the RMG sector in the post-MFA regime. While the preceding sections laid out the global and domestic context inwhich the BangladeshRMGindustrymust now operate inlight o fthe absence o f quotas, inthis section, we explore a number o f strategic options open to the sector for coping with the challenges faced under the new global regime and seizing the opportunities presented. Some strategic options open to the garment sector could be categorized as follows: Policies to enhance global competitiveness: 0 minimize anti-export bias o f trade policy and remove behind-the-border constraints to overall export competitiveness (World Bank, 2005a); 0 raise RMGproductivity by courting FDI. Policies to reduce time to market: 0 reduce lead time; 0 address the issue o fbackward linkages; 0 explore CBW option. Takingadvantageof externalpolicy environment: 0 seize opportunities for preferentialmarket access; 0 take advantage o fregional cumulation; 0 negotiate FTA with India. Strikinggovernment-industrypartnership to developforward linkages. 0 Product and market diversijkation 0 Image building and brand development 0 pe$ormance rating database 19 FDIor Joint ventures 0 Specialized units 0 Human Resource Development Explore employmentoptionsfor displacedgarment workers. Enhancinp Comvetitiveness 7.2 Export concentration in readymadegarments makes the economy,jobs and income, extremely vulnerable to external shocks arising from changes in global demand for RMG. Export diversification, therefore, is a high priority. But the import regime for other exports i s cumbersome and the duty drawback system dysfunctional, thus literally stopping any prospect o f export diversification dead on its tracks. It must be noted that streamlining the import regime for exports in a highly protected economy such as Bangladesh i s just as important for ensuring export competitiveness as reducing anti-export bias o f the tariff and QR regime. 7.3 Reduce the remaining significant anti-export bias. Although Bangladesh has managed to insulate the RMG sector through the bonded-warehouse and EPZ schemes from the effects o f protectionist policies, further trade reforms are needed to provide comparable stimulus to other industries with the potential o f diversifying the country's exports [World Bank, 2005al. Inaddition to dealing with the flaws ingovernance, transport, and finance, trade-relatedpolicychanges needto further simplifythe importtax regime and reduce the dispersion and average level o f nominal (and thus effective) protection, preferably through a pre-announced medium- and long-term schedule o f tariff reductions (as done recently by India), and the elimination o f any remaining protective quantitative restrictions. At the same time, Bangladesh should avoid using direct export subsidies because o f the strong likelihood o f their abuse and limited impact. Policy reform should focus instead on addressing the costly bottlenecks that damage not only performance but also the investment climate. 7.4 Specificallv to bring high payoffs in export competitiveness, reform should be a priority in Customs as well as theDutv Exemvtion and Drawback Office (DEDO). Practices inboththese services are characterized by inefficiency and rent-seeking that penalizes firms capable, in a freer regulatory environment, o f expanding Bangladesh's share o f world trade. 7.5 A thorough-going program is required to match infrastructure services to the needs of a growing, outward-looking economy. The basic goal should be to redefine the role o f the public sector from service provision to regulation while encouraging private-sector participation by establishing rules of competition, strengthening the regulatory environment, and addressing the problem o f poorly performing state-owned enterprises (SOEs). Specifically, it i s important to remedy the power and transportation infrastructure failures that exporters see as heavy burdens on their competitive performance. While there i s no doubt that additional investment i s required in energy and transport infrastructure, the World Bank (2005a) report concluded that the changes that are required most immediately are modifications ofpolicies, processes, and management rather than capital investment 7.6 Raisingproductivity is key to competitiveness. The broad aim o f policy measures to improve Bangladesh's competitive performance i s to expand exports and to strengthen growth prospects by raising factor productivity and encouraging private investment. Total factor productivity (TFP), i.e., the efficiency with which resources are used in production, has been crucial at the margin to the country's overall economic growth for nearly three decades. Clearly, it i s important to improve human capital and to mobilize more investment from foreign and domestic sources. What i s also critical for export- accelerated growth, however, i s the efficiency with which labor and capital resources do their work. Inefficiencies o f various kinds and degrees o f severity now hamper all sorts o f enterprises. 20 Box 4: Key Recommendations of the Growth and Competitiveness Report Trade policy: Unify andmerge allpara-tariffs with the CD. Reduce maximum tariff to 15 percent or lower over the next 5 years or so. Replace all trade-related QRs with appropriate tariffs. Simplify and automate `duty drawback system'. Overhaul DEDO Strengthen commercial sections o f Embassies, with performance-based appointment and promotion o f commercial counselors. Fiscal issues: Phase out export subsidy, including some new ones added inthe past 2-3 years. Continue general reductions in CDs o n machineries and machinery spares, with rates on both to be commensurate. Regulatory processes and governance: Simplify procedures for registration o f firms, issuance o f licenses, and abolish unnecessary licenses. Streamline inspection regime. Streamline import-export regime for other exports; Improve customs administration. StrengthenASYCUDA; Introduce Direct Trader Input System. Labor: Modify labor policy to allow night shifi work for women. Provide adequate dormitory facilities for workers. Establish good system o f reporting o n compliance standards. Finance: Strengthenfinancial sector reforms to bringabout lower interest rates. Infrastructure and utilities: Invest inport infrastructure and container terminals. Install ship-to-shore gantry cranes. Grantpermission to establishoff-dock yards to handle inbound containers. Permit sealed containers/ vans to move across the border. Introduce metered gas with necessary investment ingas pipelines and infrastructure. Widen roads and increase capacity o f Dhaka-Chittagong road corridor. Adjust prices o f freight services and increase number o f freight trains. Introduce commercial management inRailways. Quality, social and environmental standards: Establish good quality testing laboratories and increase existing laboratory capabilities. Improve Govt. and industrycapabilities inhygiene and other food safety controls. Invest inharbor/ landing facilities, ice-water supplies, etc. for shrimpproduction. Implement the "Shrimp Seal o f Quality" program. Establishand enforce environmental standards. ExtendHACCP to farms. Help farms to upgrade production methods. Enterprise capacity and marketing: Develop information services inEPB, BGMEA a n d or BIFT. Support trade missions, participationintrade fairs, buyer-seller match makingand training. Develop financing mechanisms for sustainable training (esp. contribution from industry). 7.7 The link between the two - productivity and competitiveness - is too close to be ignored. Productivity growth i s explained by higher output with given level o f inputs, or a situation where the same level of output can be had from a smaller combination o f inputs. A host o f factors could be driving productivity growth: improved management, labor force with better skills, improved technology, higher quality inputs, improved infrastructure logistics. Productivity improvements enhance competitiveness; 21 lack o f it erodes competitiveness. Most analysts have opined that relatively low productivity in Bangladesh RMG sector undermines its competitive strength stemming from low wages. 7.8 I n developing countries such as Bangladesh, productivity change is ojlen constrained by, among other things, lack of capital, technology and management. Foreign direct investment (FDI), by bringingthese scarce inputs as well as market access, could be a critical driver o f productivity change. Policies and practice suggests that Bangladesh has been a reluctant receiver o f FDI at least in the RMG sector as, until recently, FDIin RMG firms was limited to EPZ only while the domestic sector was kept o f f limits for foreign equity or partnerships. Unlike many developing countries such as Cambodia, Mauritius and Mongolia, where most o f the garment firms are part o f some larger multinational corporations in the form o f foreign direct investment, less than 15 percent o f Bangladesh garment firms have foreign equity. This i s partly due to the restrictive policy condition which required that any FDIin RMG be associated with simultaneous investment in a backward linkage facility (spinning, weaving, dyeing and finishing). The ostensible reason for this restriction was to safeguard rents from quota allocations to Bangladeshi entrepreneurs. We believe this policy could have been responsible for insufficient productivity improvements inthe RMG sector - something that would have stood Bangladesh ingoodstead inthe post-MFA competitiveenvironment. 7.9 The restriction on FDI was finally removed in the Industrial Policy 2005; but perhaps the Bangladesh RMG sector missed out on certain benefits that accme from the presence o f FDIin a sector. Conventionalwisdom has it that firms with foreign equity tend to be more productive. This could be due to the firm specific tangible assets such as exclusive technology and product designs, or the intangible know-how embodied in foreign equity such as superior management, marketing, networkmg and sourcing. Such assets may be more readily available inbig multinational corporations (MNC). As such, being part o f MNCs allow the local subsidiaries with foreign equity to gain access to these assets, which inturn make them produce more output given the same level of inputs, and thus a higher level o f total factor productivity (TFP) than the solely domestic owned firms. Such hypothesis has some empirical support based on samples o f Venezuela manufacturing firms studied inAiken and Harrison (AER, 1999) and Malaysia service sector firms inKee (forthcoming). 7.10 To test the hypothesis that FDIwas beneficial to the RMG sector, we carried out a survey o f 350 RMG firms, including 49 firms with foreign equity located in the EPZ. The study based on this survey first examined the productivity changes in the RMG sector, and then looked at the potential productivity advantage o f FDI firms operating in Bangladesh. In addition, the study tried to identify the possible channels by which local firms could benefit from the FDI firms - the productivity spillover effects, beyond the physical presence o f FDIfirms. 7.11 The main focus of the study was on estimatingfirm productivity - measured by total factor productivity, which i s the level o f output that is not explained by inputs -- by modifying the state o f the art technique due to Olley and Pakes (Econometrica, 1996), to control for firm and year specific biases. Comparing firm productivity across all firms in all sub-industries and locations yields some interesting insights in terms o f relative productivity of firms as well as productivity change. A comparison o f different firms across different sub-industries, on average, reveals that knitwear firms are the most productive. An average knitwear firm has 10 percent higher productivity than a woven firm, and i s 17 percent more productive than a sweater firm. Figure 1 presents the distribution o f the estimated firm productivity by the three sub-industries. Interms o f locations, productivity o f firms located in Dhaka- EPZ is the highest (FDIeffect), followed by firms inDhaka, Chittagong-EPZ and Chittagong. Figure 2 presents the distribution o f firmproductivity by location. 7.12 Comparing firm productivity from year to year within firms also sheds some interesting new light. On average, garment firms are 3 percent more productive in2003 than in 1999. The improvement 22 in productivity is especially clear for the sample of domestic firms -- on average, domestic firm productivity i s 5.5 percent higher in 2003 than in 1999. Figure 3 presents the movement of firm productivity over time inthe different sub-industries. It i s clear that most o f the improvements are driven by firms inthe Sweater andWoven garment industries. Thus we findthat although knitwear firms are on average more productive, sweater and woven garment firms are fast improving their productivity. These results purely reflect the growth inproductivity within a given firm, and thus are not contaminated by the composition o f firms in different industries. The evidence suggests that although there i s widespread perception o f low productivity o f Bangladeshi FWG firms, there i s positive change over the short period picked up by the survey. Such an increase in productivity within a firm suggests that there are some exogenous factors pushing firms to be more productive over time. We explored one such exogenous factor which i s the productivity spillover effects o f FDIfirms. 7.13 Preliminary Findings Based on Firm Survey. Firm level survey was conducted from the period of Figure 1: Distribution of Firm Productivity by Sub-industries November 2004 to April 2005, covering a stratified random sample o f 350 firms, which i s about 10% o f 1.7 _I L 2.7 the total population o f the garment firms currently 2 . 8 operating inBangladesh. After cleaning up the data to t:* - 2 . 6 D) Q 2 . 5 exclude outliers and firms with incomplete 2.5 information, there are a total o f 231 firms in the 2.4 2 . 4 unbalanced final panel data set o f 1026, from 1999 to Knitwear Sweater Woven 2003. In this unbalanced panel data set, the composition o f sub-industries o f knitwear, sweaters and woven is 24%, 8% and 68% respectively. Among Figure 2: Distribution of Firm Productivity by Location the sampled firms, 13% have foreign equity, while the 2.727 remaining 87% are purely domestic owned. 2.7 2.68 2.66 7.14 Table 11presents the sample means o f the key L i2.64 variables o f the sub-industries o f knitwear, sweaters 5 2.62 and woven, by foreign versus domestic firms. It is 2.6 clear that FDI firms are in general larger in sales and 2.58 2.56 exports; they purchase more material inputs, including 2.54 imported materials; they hire more employees, 2.52 including production workers. FDI firms also have ChlttaQOnQ ChittaQOnQ- - - Dhaka Dhaka-EPZ Others EPZ larger capital stock and investment. All these suggest that FDIfirms are larger in scale and presumably more profitable and productive. To formally study the Figure 3: Productivity Growth of Domestic Firms productivity superiority o f FDI firms, and the possible 2.75 1 by sub-industries (YO) productivity spillover to domestic firms, we first 2.7 needed to estimate firm level productivity for the firm 2.65 sample. The estimated firm productivity was then 2.6 related to the ownership o f the firms, and the relationshipbetween productivity o f domestic and FDI 2.55 firms in the same sub-industries were statistically 2.5 examined. 2.45 2.4 7.15 Are FDI firms more productive? To 1999 2000 2001 2002 2003 formally study the overall productivity o f firms, we Knitwear Sweater E! Woven needed to estimate firm production functions, taking into account total factor usage per unit o f output. Tostudy the effects o f FDIon productivity, firm level productivity was then related to ownership characteristics of the firms. This was done by relating 23 Figure 4: Productivityof Firms with Different Ownerships estimated productivity to the ownership structure o f firms using between firm panel regression, controlling for industry, year, and location fixed effects. The results show that firms with foreign equity are on average 20 percent more productive than otherwise identical domestic firms. The productivity advantage of FDI firms is robust to age and export destinations. In addition, relating the productivity performance o f Domestic FDI Firms domestic firms to that o f FDI firms we found that there firms were indeed positive and significant productivity spillovers. For every 10 percent increase in the productivity level of FDIinthe industry, productivity o f Figure 5: Number of Firms in Different Markets domestic firms increases by 1.4 percent. As shown in Figure4, on average, productivity of firms with foreign Distribution of Firms by Markets equity was found to be about 20 percent higher than m 2500 7 I purely domestically owned firms. 2000 IC 1500 rc 7.16 What could have explained the 20 percent 0 1000 L productivity advantage of FDI fzrms? To answer that P a 500 question, we needed to isolate the effect o f foreign E O ownership from the influences o f sub-industries, investment climate o f the locations, and the macro economic shock in each year. Given that ownership seldom change within firms in our sample, between- export destinations firms variation in foreign ownership was used to identify the effect o f F D f o n productivity. The result shows that a FDI firm is still about 20% more productive than a domestic firm in the same industry, location and year. This shows that the effect o f foreign equity on firm productivity i s independent o f the location o f the firms, the sub-industry o f the firms and the macroeconomic fluctuations. In addition, the regression results confirm that FDI firms do have higher productivity even after account i s taken of the export destinations and thus the potential demand shocks that firms might be subjected to, as well as the experience o f firms, as proxied by age. Finally, the results also confirm what the earlier finding from EPB data showed- firms exporting to USA tend to be more productive. Table 11: SummaryFeaturesof SurveyedFirmsinRMGIndustries Figures are averages inUS$ `000 Knitwear Woven Sweater Domestic FDIfirm Domestic FDIfirm Domestic FDIfirm firm firm firm Sales 3,050.89 5,044.48 2,926.72 13,900 2,363.51 3,603.47 Export 2,951.96 5,044.48 2,919.79 13,900 2.362.95 3,603.47 cost 2,9 17.29 4,195.3 8 2,s 87.39 12,700 2,141.49 3,350.96 Material 2,037.68 2,888.02 2,015.82 9,665.94 1,43 5.53 2,389.08 Imported material 1,560.67 2,569.80 1,590.77 8,393.14 564.88 1,811.85 Employees* 582.94 996.23 600.58 1,893.18 906.89 1,305.85 Prod workers* 501.52 943.70 560.29 1,790.30 859.63 1,214.75 Capital 2,033.42 1,5 10.17 639.52 5,076.09 1,002.34 4,231.34 Investment 817.58 79.3 1 57.78 315.92 215.66 344.92 (*)These numbersrepresent average employmentper fm. 7.17 Export Performance of Garment Sector andproductivity. The past few years have witnessed an expansion o f Bangladesh garment export to the world market. In 1998, the total value of garment export 24 from Bangladesh was about US3.8 billion; it increased to US$4.9 billion in 2003, US$5.7 billion in 2004, and i s expected to be $6.3 billion in 2005. Some evidence o f productivity improvements can be gauged from the distribution o f firms by choice o fmarkets. 7.18 Interms of the distribution of firms across different markets in 2004, there were 1967 firms exporting under GSP, mainly to the European market, 1039 firms exporting to the US, o f which 709 export under quota allocations, and 1231 firms exporting to other countries. Figure 5 presents the distribution o f firms by export destinations. Among these firms, 26 percent only supply to one market, 39 percent supply to two markets, 24 percent to three markets, and 11percent to all four markets. This is clearly presented inFigure 6. 7.19 Figure 7 presents the choice of export markets o f Bangladesh garment exporters according to the number o f export market the firms supply. It is very clear that EU i s the most popular destination, especially among firms that have only one export market. Among the 1109 firms that only supply one market, nearly 850 firms concentrate on EU which i s about 76 percent. The US market appears to be toughest to break inamong this group of firms, less than 8 percent only export to the U S with and without quota. For firms that supply two markets, both the EU and the others are the favorites. Together, they account for 80 percent of the markets among the 1640 firms that export to two markets. The U S inquota market i s popular for firms that export to more than 2 markets. 7.20 In addition, according to Eaton, Kortum and Kramarz (AER, 2004) who study the export performance o f French firms, the number o f markets a firm supplies reflects the productivity and Figure 6: Number of Firms vs. Number of Markets competitiveness o f the firm in the world market. Distribution of Firms by Number of Markets The above distribution o f firms implies that more than 35 percent o f Bangladesh garment exporters 1400 participate in world markets widely with at least 3 6 1200 c 1000 export destinations, and are thus very competitive. O 800 This is quite evidence inFigure 8, when we plot the 600 E 400 unit value o f garment export (left axis) and total 2 200 export value (right axis) against the number o f 0 export destinations. Firms that export to more 1 2 3 4 destinations tend to have higher average unit values number of export distinations and larger in size, with the former reflects better quality and the latter indicates greater scale Figure 7: Market Choice by Firms with Different Markets economies, both signal higher productivity of the Choice of Export Market firms. The differences in unit values and total size 900, among firms with different number o f markets are 800 statistically significant. = c 500 400 7.21 Conclusion. Overall, there is convincing 0 300 and statistically significant evidence suggesting that gs 200 FDI firms are more productive than otherwise 100 o identical domestic firms operating in Bangladesh. Thesignificant spillover effects are not to be ignored either. This suggests that the earlier policy of restricting FDI in the domestic RMG sector might have harmed it by not helping to raise productivity export destinations enough toface global competitionfrom a position of ElOremMfirms Qhromrktfirms OthreemrMfirms ~fourmrk!firm strength. Bangladesh needs to make up for lost time and ipportun7ty. Now that the FDI ie&ction has been removed from the Industrial Policy 2005, all 25 efforts should be made tofacilitate FDI inflows into the sector to take advantage of better management, technology, capital and market access. Ovtions for Reducing Time toMarket 7.22 Reducing Lead Time is critical. Lead time - also known as turnaround time -- refers to the time it takes from the receipt o f a confirmed order or L C to the delivery o f the product to the buyer. It has emerged as an important issue due to certain structural changes inthe global market. Many of the changes that have occurred (or one in the offing) in the world apparel industry were driven by exigencies of the competitive retail market where firms have to be lean and mean inorder to survive or thrive. The apparel market i s no longer bi-seasonal as in the old days. Frequent fashion changes and consequent rapid obsolescence have made it multi-seasonal. The shelf life o f fashion clothes and high-value items i s accordingly much reduced. The retailers look for flexibility inthe supply chain such that products can be puton shelf as and when demanded by consumers. The drive for reductioninleadtime is thus an attempt to conform to the market demand as well as achieving greater supply chain efficiency. As observed by Lezama et a1 (2004), the inexorable pressure for reduction in lead time has grown and the supply chain had to respond by reducing the lead time. A major buyingcompany reduced their average lead time from 21 weeks in 2000, to 18 weeks in 2001, to 14 weeks in 2002, and to 12 weeks in 2003 (p. 36). The progressive reduction i s noteworthy as it conforms to the supply chainhypothesismentioned above. 7.23 Another developmentthat callsfor shorter lead time is the recourse to replenishment orders by retailers with the advent o f lean retailing. One of the important recent changes in retail inventory management has been the introduction o f vendor managed inventory or continuous replenishment programs which entail determination o f the optimum stocking levels at the retail point, and a steady transmission o f current sales and stock data to the manufacturers in order that they can replenish the stocks to maintain them at their desired levels continu~usly.~~This mightrequire quick response fromthe manufacturers. Replenishment orders are more prevalent for the basics and fashion basics as their demand are more stable. This tendency could also create pressure on the RMGmanufacturers o f Bangladesh, who producemostly basics and fashion basics, for shorter lead time. 7.24 Lead time and size of the order also vary a great deal depending on the type of the product. Fashion retailers need to carry inventories of products o f varying styles and for shorter duration. They operate on smaller but more numerous orders and shorter lead time. Retailers o f basic garments, on the other hand, tend towards larger orders, but fewer items and longer lead times. Since the styles o f basic items do not change frequently and demand for their products i s more stable, mass market retailers can be Figure 8: Exporting and Productivity more flexible about lead times. Notwithstanding this UnitValue, Total Exportby Number of Markets attribute o f the mass market, it should be borne in 50 10 - mindthat, other things remaining the same, a retailer " e 2 9 40 i s likely to prefer shorter to longer lead times. This .E 30 J wouldimplyvendors with longer leadtimes will have 6 5 to provide greater incentives or more favourable 7 2o F 4 ' ' 3. terms in other areas, such as price and quality, in = 10 2 -c order to offset the attractiveness o f shorter lead time 0 0 : to the buyers. This cuts into the profit margins o f k2 vendors with a longer lead time and reduces their number of export destinations competitive strength., unit value per dozen +total value of export 25 See Abemathy et al(l999)for an excellent discussion. 26 7.25 The importance of timefactor infuture RMG trade has been recognised also by OECD (2004). The study finds that lead time will play a crucial role indetermining international competitiveness. "Time factors can be an important trade bamer for intermediary inputs involved inan internationally fragmented production process. There are trade-offs between low wage cost and time factors since temporal proximity to large consumer market provides a competitive edge in the highly competitive, time sensitive and fashion-oriented clothing market" (p. 25) Thus, OECD warns that countries that rely on imported intermediate inputs (such as fabric) and have a long lead time will face some competitive disadvantage in the post-MFA world. Lead time was also mentioned by leadingbuyers o f apparel as one o f their two main concerns ina discussion meeting with the stakeholders inDhaka recently. Table 12: Leadtime for servicing an order for RMG Lead time (days) Cambodia 90-120 China 45-60 Indonesia 60-90 Malaysia 60-90 Thailand 60-90 Vietnam 60-90 Bangladesh 90-120 Source: BGMEA (2004a) 7.26 Bangladesh is in the unenviable position of having the longest lead time among its major competitors(Table 9). Its lead time varies between 90-120 days (for woven garments). It i s considerably shorter for knitwear items (45-60 days) since most raw materials (yam and knitted fabric) are procured domestically. The breakdown o f the total lead time into its various components i s shown in Table 10. This breakdowni s very revealing as it indicates the underlying reasons for the longer leadtime for woven exports. 7.27 I t takes a factory 42-60 days to receive the raw materials (essentially fabric) at its premises from the time it receives a confirmed LC. Several factors contribute to this long time required to receive the raw materials. First, most o f the fabric requirements of woven RMG (80 per cent or more) are imported from such distant countries as China, Korea and Taiwan (as well as less distant India). Given the geographic position of Bangladesh and the prevailing shipping routes, it takes considerable time for shipments from these countries to reach the ports in Bangladesh. The time requirement i s further lengthened by the fact that goods have to be brought from China and elsewhere through transshipment at Singapore or Malaysia as goods cannot be shipped directly to Bangladesh fi-om these countries. As a result it takes 25-30 days for the consignment to reach local ports. Archaic port facilities, andbureaucratic hassles add to the time it takes to take delivery o f the consignments. Bangladesh i s situated farthest away from its main market U S A and also very considerable shipping distance from the European countries. Further, the export consignments have to be sent by feeder vessels to Singapore or Malaysia to be loaded into mother vessels to carry them to their final destinations in U S A and Europe. Consequently, it takes about 28-30 days for export shipment. The time i s reduced by about a week or so ifmore expensive faster vessels are used for shipment. 7.28 I t is evident that much of the excess time requiredfor processing an order is due to having to procure the raw materialsfrom overseas.If raw materials could be sourced locally the time required the to receive them at the factory gate could be reduced from 42-60 days to just one or two weeks. Hence, local sourcing could reduce the lead time to 55-75 days. This would make the lead time o f Bangladeshi garment exporters comparable to that o f its major competitors. 27 Table 13: Typical lead time componentsfor Bangladesh(days) Days taken after ceding step Optimal Non-optimal Exporter receives c o n f i i e d L C 0 0 Raw material supplier receives LC 4 6 Raw material supplier produces and ships goods 7 15 Ship berths at port o f import 26 30 Raw materiali s unloaded and taken delivery fromport 4 7 Consignment reaches factory 1 2 Garment packed and shipped 20 30 Consignment reaches buyer's port 28 30 Total 90 120 Source: BGMEA (2004a) 7.29 The discussion above clearly suggests that thefirst best optionfor reducing lead time is to have the raw materials of RMG industry produced domestically. This has already happened in varying degrees for different raw materials. Duringthe last decade and half Bangladesh has become almost self- sufficient inthe production o f many accessories items. Itnow produces, according to BKMEA leadership, 90 per cent o f the knit fabric requirements o f the export-oriented knit garment manufacturers. This enabled the sector to reduce the average lead time to only 50-60 days. However, despite all the incentives and large protective barriers, the woven fabric section o f the textile sector has not been able to reduce the demand-supply gap. It can meet only about one-fifth o f the total requirement o f the woven Rh4G sector, and this situation has not changedmuch for the last five years. The excess demand for woven fabric (over and above domestic supply) has, therefore, increased over the years. The unavailability o f domestic woven fabric not only increases lead time, but also causes GSP related disadvantages as discussedlater. It poses a serious constraint to export competitiveness. 7.30 The next section on backward linkages discusses the reasons for insufficient investment in the woven fabric production facilities. It seems fairly certain that the demand-supply gap for woven fabric cannot be reducedinthe short to medium term; in fact it may increase. Consequently, lead time cannot be reduced by domestic sourcing o f fabric alone. 7.31 Although it may not be possible to increase domestic supply o f fabric at par or in excess of the increase in demand within a short period, it might be possible to meet part o f the demand by increasing the capacity o f dyeing and finishing industries. If these industries are allowed to stock up grey fabric in advance o f orders, they could supply finished fabric to the woven garment manufacturers at fairly short notice. The lead time would not be greater than when domestic fabric i s used. Hence, this option should receive serious consideration. 7.32 This option also cannot be expected to dramatically change the demand-supply gap within a few years as the gestation time would be long; but the next few years will be crucial for the health o f RMG woven industries o f Bangladesh. Ifthey progressively lose their market due to their inability to supply the market at short notice due to long lead time, they will find it difficult to recapture it in the future since markets once lost are not easily or quickly regained. With the loss o f market, demand for fabric will decline, reducingthe total market o f the local textile industry. Ironically, inthis event there will be greater self-sufficiency, i.e. a reduction inthe demand-supply gap; but this will be achieved through a reduction indemand, not anincrease insupply. BothRMGand the textile industries standto lose insuch a milieu. They will not be able to contribute to national development and poverty reduction as robustly as envisaged. Over-protectiveness o f one industry could end up injuring several other industries and the economy as a whole. 28 7.33 A third option that has been mootedis building up a close relationship with nearby suppliersin order to receivefabric at short notice. Currently the RMG manufacturers are free to import fabric from these countries. But they have sourced only a small fraction o f their requirement from nearby countries such as India, Thailand and Palustan. Interviews with producers reveal that bulk o f the fabric i s currently sourced from China, Hong Kong or S. Korea. Long-term relationships with suppliers, buyer preferences for fabric sourcing, dependability regarding timeliness and quality, are some o f the determinants o f such distant sourcing rather than from the neighborhood. This would suggest that, for woven garments, market conditions or other constraints have thus far restricted imports o f fabric from the region, --a process which could have cut lead time substantially. . 7.34 I t appears that certain policy constraints have prevented Bangladeshfrom utilising available opportunities to reduce lead time by sourcing inputsfrom the region. There i s also some reluctance in forging closer links with India or Pakistan for supply o f inputs at short notice. Many integratedknit units could reduce leadtime by importing yam from India by landroutes. However, the government bannedthe import o f yam by land routez6.Importation by sea routes adds to transport costs and often takes 10-20 days extra to reach Chittagong port, thus removing much o f the advantage o f obtaining yam from India. The limited opportunity that was available for reducing lead time to meet tighter delivery schedule was eliminated by this ban. It i s difficult to imagme any good reason for imposing ban on import by land routes. It i s hurtful to local industries, bad for bilateral relations, and adversely impacts on exports and balance o fpayments. The case for its withdrawal appears compelling. Recommendation: Overhaul Chittagong port facilities, improve infiastructure and transport logistics, and streamline the import regime by removing cumbersomeprocedures. End the ban on textile imports through landports. 7.35 Backward Linkages: what are the options?. The ready-made garments industry grew up in Bangladesh without the benefit o f either backward (upstream) or forward (downstream) linkage industries. During the early years, the RMG industry in Bangladesh purchased almost all its material inputs such as accessories and fabric from overseas. This was also the period of most rapid growth o f RMGexport. As the productionof RMGincreasedby leaps andbounds, so didthe import o fthese inputs. Local entrepreneurs perceived profit-earning opportunities in the manufacture o f these inputs. Since the investment cost o f manufacturing accessories i s relatively low, this was the first area exploited by the domestic entrepreneurs. The local accessories industry now supplies most o f the requirements o f the RMGindustryand is also exporting small quantities. Table 14: Consumptionofwoven andknitfabricby export-orientedRMGsector Year Woven fabric Knitfabric RMGTotal Domestic Total Domestic Total Domestic Total DomesticlTotal 1997-98 102 1215 230 585 332 1800 0.184 1998-99 132 1200 289 660 421 1860 0.226 1999-00 150 1234 448 815 598 2049 0.292 2000-01 190 1317 565 943 755 2260 0.334 2001-02 200 1386 741 1140 941 2526 0.373 2002103 224 1491 1046 1245 1270 2736 0.464 2003104 261 1629 1484 1649 1745 3278 0.532 Source: BDXDP (2001) and Gherzi et aZ(2002) and staff calculations. 26 The rationale for ths banwas to prevent smuggling o fyam over landroute where monitoring at customs border stations was slack. Inreality, this measure gave added protection to textiles while raising input costs for RMG exporters. 29 7.36 The increase in import of fabric also encouraged domestic entrepreneurs to move into the primary textile sector,which was then dominatedby loss-making public enterprises. The sale o f domestic fabric to the export-oriented RMG industryincreased very rapidly from only 45 million square metres in 1993-94 to more than a billion square metres by 2002-03 (BDXDP2001 and Table 1l).27helped This was by generous cash incentives given to textile production for exports or deemed export and the duty-free access to EUmarket. 7.37 However, the two sections of RMG, knit and woven, did not beneJit equallyfrom the expansion of domesticfabric production. Most o f the fabric sold to the RMG industry was actually usedby the knit section o f the industry. As shown in Table 11nearly 80 per cent o f the total sale o f fabric to RMG was made to the knit section by 2001-02. The Bangladesh Knitwear Manufacturers and Exporters Association (BKMEA) leadership claimed during a discussion that the production of knit fabric has expanded sufficiently to meet most o f the demand for such fabric (more than 90 per cent) by the knit industry.28In contrast, woven garment industry i s still overwhelmingly dependent on imported woven fabric. Domestic producers can meet less than one-fifth o f the total demand o f the woven section, which is by far the larger section o fthe RMGindustryaccounting for about 60% o fthe total value o f exports (and output) o fRMG. 7.38 Three factors appear to have been responsiblefor this uneven growth in the utilisation of domesticfabric by the two sections of the RMG industry. First, most o f the knit garmenting i s done in integratedplants which house the major part o f the country's knitted fabric processing capacity (Gherzi et aZ2002). The growth o f such plants has been facilitated by the low cost o f investment and the relatively simple skills required to operate them. The integratedplants can knit fabric from yarns. Spinning industry also grew with the increase in the demand for yam. Local spinners now meet about 60-70% o f the demand o f the knit section o f RMG for yam. The rest i s imported. The knit fabric i s dyed,and processed inthe same establishment for use as inputsfor knitgarmenting. 7.39 Second, the real boost to knit garment manufacturing was given by the change in the applicable rules of originfor availing European GSPfrom three to two stages. The relaxation o f the stringency o f the rules o f origin opened up a huge opportunity for the knit garment manufacturers. Many of them invested inintegrated knit plants as the cost o f investment was relatively modest. Knit exports to EUnow qualified for duty-free and quota-free treatment as most o fthe knit fabric usedbyRMGindustry was manufactured locally. This gave knit exports from Bangladesh considerable price advantage over non-LDC competitors since the MFN duty rates on garments in EU were relatively high (up to 12.8 per cent).29 7.40 The thirdfactor promoting the rapid growth of knitfabric manufacturing was the granting of generous cash incentives in 1993-94 on exports of (garments madefrom) fabric produced locallyfrom domestically spun yarns. This encouraged large investment in spinning and composite textile mills despite the fact that the costs o f production o f yarn and fabric were considerably above international prices and there was surplus capacity inthe international market (Table 15). There was some skepticism ifthe domestic textile industrywould remaincompetitive once the cash incentives are phasedout inJuly 2005. However, the reduction of the cash incentive from 30 per cent a few years ago to only 5 per cent now must have reduced the effective price received by the spinners by a large amount; but this has not reduced the supply o f domestic yarn. In any case the cessation o f the cash incentives will remove a support mechanism and force the spinners to stand on their own. It should be pointed out that the payment o f cash subsidies to exports contingent on the use o f domestic over imported goods (inputs) i s *' Mucho f this i s actually self-sale. 28 BGMEAleadership concur withthe claim. 29 For illustration of this price advantage, see section on preferential access discussed later. 30 inconsistent with Article 3.1(b) o f the Agreement on Subsidies and CountervailingMeasures o f the WTO Agreement. Table 15: Relativepricesof typicalknittingyarns (US$/kg) Y a m count N e Market prices inBangladesh o f yam Bangladeshi yam Indian yam Pakistani yam Local market RMG Carded 2011 1.60-1.70 1.75-1.85 1SO-1.65 1.55-1.70 Carded 2411- 2611 1.65-1.75 1.95-2.10 1.55-1.75 1.60-1.SO Carded 3011 1.80-1.95 2.05-2.20 1.70-1.80 1.75-1.85 Combed 2011 2.15-2.40 1.75-1.95 1.80-2.00 Combed 2411-2611 2.10-2.30 2.25-2.50 1.85-2.20 1.90-2.25 Combed 3011 2.25-2.4 2.35-2.60 2.00-2.30 2.05-2.30 Source: Gherzi et al(2002) 7.41 A similar development did not happen in the upstream industries of the woven section of the garment industry. The principal reason for this is perhaps the large cost o f investment required to set up upstream linkage industries such as spinning and weaving.30 While the relatively modest cost o f investment for integrated knit plants enabled the knit garment manufacturers to move to knit fabric manufacturing, the same was not possible for most woven garment manufacturers. According to the estimates shown in Table 10 (p.14), the investment cost o f manufacturing capacity o f certain volume of processed knitted fabric i s less than one-third o f the cost o f creating the same capacity for processed woven fabric. Hence, unlike the knit industry, woven fabric manufacturing and woven garment manufacturing remained separate (nonintegrated) activities. The large investment requirements o f primary textile manufacturing and low international prices due to worldwide excess capacity limited the growth o f woven fabric manufacturing. It has barely managed to retain its share o f the demand o f woven FWG for woven fabric at around 15 per cent during the last several years. The trend o f consumption o f domestic fabric by the woven sub-sector seems to suggest that the domestic primary textile sector will be unable to supply the greater part o f the total demand for fabric o f woven RMG in the near future. However, substantial FDI in this sector could change the picture fairly rapidly. There are some encouraging signs that foreign textile manufacturers are showing keen interest in establishing production facilities in Bangladesh. The government and the private sector should put in effort to ensure that this interest i s turned into actual investment. FDI will also help the overall productivity o f the sector by bringing in advanced technical know how, better management practices and wider linkages with the international market. 7.42 Can there be substantial expansion of domestic capacityfor woven textile in the near future in light of global excess capacity in yarn andfabric? That depends on whether production costs o f yam and fabric at home are competitive with international prices. If cash subsidy and other protection become absolute necessities to enable domestic industries to continue operating, it could have negative implications for the downstream garment industry. In the past these industries were provided with generous cash incentives (30 per cent o f output cost) and total prohibition o f imports o f competing products. Although the cash subsidy has been reducedto 5 per cent (scheduled to be phased out by July 1, 2005) and import bans have been withdrawn, the sector i s still protectedby very highimport taxes3*.The protection given to the spinning and weaving sectors has some spill over effects on the RMG industry. The cost o f procuring yam and fabric domestically i s considerably higher than the international prices leading to higher costs o f production o f RMG products that reduces the competitiveness (see Table 15). 30 These activities also require highly skilledmanpower and specialized knowledge that the country lack. 31 Protective duties on fabrics and yam are 72% and 39%, respectively. 31 Imports under bond being free, the highprice o f domestic inputs apparently should not matter. However, import from overseas adds about 5-8 per cent to the cost o f input on account o f transport and port clearance charges. More, importantly, it takes about 30-60 days longer to procure inputs from overseas. Thus, Bangladesh RMG producers suffer from a natural cost and time disadvantage against producers in countries who can access inputs domestically such as in China, India and Pakistan, or RMG producers who are in close proximity to major producers o f the inputs and are not subjected to selective import restrictions. 7.43 The immediate upstream activity in the PTS sector for the RMG industry is fabric processing or dyeing andfinishing. As such this section o f PTS has important influence on the performance of the RMGindustry.The knittingsub-sector cannow meet most o f the demand for finished knit fabric, butthe weaving sector cannot satisfy more than a small fraction o f the total woven demand (see Table 11). The BTMA reports that the current capacity is over one billion metres, up from some 650 million meters in 1996-97. Though breakdown o f the capacity between knit and woven is unavailable, from the known fact that most o f the knit fabric i s processed locally it would seem that the major part o f the total capacity i s devoted to the knitting industry. Consequently the capacity for processing woven grey must be quite limited and much of it housed incomposite textile mills.The number o f mechanised dyeing and finishing units increased from 75 in 1996-97 to only 115 currently. There has not been any significant investment in the semi-mechanised units.This lack of large investment in this activity is surprising since both the major studies in the textile sector done duringthe last half a dozen years (Dr Martelli 1999 and Gherzi et a1 2002) recommended more emphasis on fabric processing. Dr Martelli Associates had concluded that "conditions for setting up new fabric processing operations on a converter basis appears to be significantly more favourable compared to setting up weaving and spinning plants.. ."(p.12). There were a number o f reasons for this conclusion. First, while there was surplus spinning and weaving capacity in the competing countries, there was a dearth o f good processing facilities in these countries. Hence, the competition in processing would be less severe than that in weaving and spinning. Second, there was a huge unmet demand for finished fabric in the domestic market. Third, fabric processing on a converter basis could reduce lead time by 6-8 weeks if sufficient grey fabric were heldin stock. Fourth, the use o f imported grey fabric would increase value addition in garment manufacturing and thereby help availing duty-free facility in EUmarket. The reduction inlead time, opportunity to rectify faults in fabric quickly and increasedvalue addition would encourage the RMGmanufacturers to source fabric locally (p. 13).32 7.44 Dr. Martelli Associates (1999) indicated that the successful operation of the dyeing and finishing units would critically depend on the availability of greyfabric. Large stocks o f grey fabric o f the most common constructions would, therefore, need to be held to supply the processing units as and when demanded. The study did not mention how this could be done, but it i s clear that measures other than the existing individual bonded warehouse system, where fabric can be imported only against confirmed letter o f credit, will have to be devised to make large scale fabric processing worthwhile for investors. One such system i s Central BondedWarehouse which i s discussed later. 7.45 I s there a case for local sourcinghackward linkage? Since an internationally competitive domestic primary textile sector contribute to the competitive strength o f local RMGby reducing cost and lead time, a favourable environment for their growth should be ensured by removing any constraints. However, this should not mean adopting (or not adopting) such measures that would raise the cost o f inputs for the RMG industryor prevent them from reducing the lead time. Industry assistance could take the form o f availability of long term loans at reasonable rates through reforms o f the financial sector, the development o f more efficient infrastructure, greater emphasis on skill development through education 32 Industry sources suggest that although investment requirements in dyeing and finishing are not very high, the skill requirements for successful operation of the industry are very specialized and difficult to acquire. This might be a reasonfor the slow growth of this sub-sector. 32 and training, improved governance that reduces transaction costs and provides security of contracts and measures to aggressively court FDI. Since borrowing cost in Bangladesh i s much higher than that inthe competing countries, some intervention might be warranted in this area. India might have used a innovative scheme (Technology Upgradation Fund) to good advantage.33 Any intervention, however, should not be in the form o f policies that compel RMG manufacturers to purchase domestic inputs at a higher cost than their international competitors, such as banning o f overland import ofyam or preventing them from the opportunityto access a ready supply o f fabric. 7.46 Thefundamental force behind the rapid growth of yarn and fabric production in the 1990's has been the spectacular growth of RMG output and export, and not textile protectionper se.34 Any dampening o f the growth of RMG export will adversely impact on the growth o f the primary textile sector. Protection gwen by restricting trade may allow the existing producers to increase their profits, but it will not provide space for further expansion o f the primary textile sector unless the sector becomes internationally competitive. If the sector attains international competitiveness, it does not need trade protection nor should such protection be given that would stifle the RMG sector and hurt PTS in the longer term. High protection o f the textile sector i s a drag on the RMG industry requiring cumbersome procedures (e.g. SBW) for duty-free sourcing o f textile inputs.Reduction o f textile protection over time is a policy imperative for enhancing competitiveness and growth o f RMGexports. 7.47 Given the relative importance of RMG in the national economic lve, it will be unwise to tie its fortunes to thefortunes of fabric manufacturing, or for that matter, any other industry. As mentioned earlier, it i s not desirable to make the growth o f any industry contingent on the growth o f another industry.It i s folly to think that a country could always have comparative advantage inall activities inthe value chain o f a product. Many textile producers and exporters do not have a comparative advantage in the production of apparel, just as many RMG exporters do not have a comparative advantage in the production o f textiles. Hence, the growth o f each sub-sector should be determined by its comparative advantage. This i s particularly true o f export industries since they have to compete in the global marketplace over which the country has virtually no control. Inorder to survive and compete inthe global market, domestic export industries should be able to access inputs from the cheapest sources, local or foreign, and exploit any (and all) advantages that might be available (such as GSP). It i s o f supreme importance to gain an edge over the international competitors, and no profitable opportunities should be allowed to pass by. The growth of the RMG industry will create opportunities for both backward and forward linkage activities, some more profitable than others. It will also create opportunities for expansion o f many co-operating activities such as banlung, insurance and transport. Unless the government or powerful sections o f the society erect insurmountable barriers, dynamic entrepreneurs will exploit these opportunities to buildup nascent as well as hitherto non-existent industries. The nation has already witnessed the remarkable growth of apparel accessories industries as well as industrial laundnes to service the RMG industries. The previously non-existent local accessories industry now supplies almost the entire demand o f the RMG sector. Recommendation: The fortunes of the textile industry must not be tied to the RMG industry. If the Government desires to provide support to the textile industry, such as interest rate or price support, it 33 Government of India established a Technology Upgradation Fund(TUF) to support investment for upgradation o f textile machinery. Textile fm borrowed from commercial banks at lower-than-market rates, the difference being re-financed from TUF. 34 Textile had extremely high protection inthe earlier years, but that did not lead to an expansion of the primary textile industry nor did it entice many entrepreneurs to invest in the industry. It is only when the growth of RMG industry had created a substantial demand for PTS products by the early 1990s that the upstream industries flourished. 33 must do so through its normal budgetary instruments in a transparent manner. The cost of support must not be shifted to the RMG industry that is totally unprotected in the international market. 7.48 Central Bonded Warehouse. An innovative scheme suggested by some quarters including BGMEA as a solution to the problem o f long lead time is the establishment o f central bonded warehouses. Conceptually CBW i s similar to the individual bonded warehouse that each export-oriented RMG factory i s permitted to have to stock duty-free imported inputs.The difference is that CBW can be set up by any firm and the imports o f duty-free RMG andtextile raw materials will not be conditionedon master LCs. The CBW operator could be permitted to stock up a whole range o f T&C inputs such as finished and grey fabric, accessories, dyes and chemicals, yam, RMG and textile machinery and spare parts in amounts determined by expected demand. The RMG and textile manufacturers can then purchase these inputsduty-free from the CBW directly. Some appropriate system will have to be put inplace (such as sale against back-to-back LCs and locating the CBW in Export Processing Zones) to ensure that the duty-free inputs are actually used in the production o f goods that are exported.35It is also possible that in order to protect the interests o f local competing industries some restrictions may be placed on the CBW regarding the range, amounts and types o f inputs it can store. 7.49 The principal difference between a firm's bondedwarehouse and C B W i s that while an individual bondedwarehouse can import inputsduty-free only against a confirmedback-to-back LC or export order, the CBW can engage induty-free imports o flistedraw materials without any back-to-back LC or export order. Thus the lead time in the case o f individual bonded warehouse for woven garments will be o f the normal duration (90-120 days) since inputs can be ordered from overseas only after an export order i s received. But inputs can be purchased duty-free directly from CBW as soon as an export order i s received. Hence, RMG exporters can save on the shipping time required for importing inputs. The lead time i s accordingly reduced by 35-45 days. Ina way, CBW replicates the advantages o f domestic supply. It shouldnot take more time to procureraw materials from CBW than it would take to procurethem from domestic producers. 7.50 Another important difference i s the scale o f operation. The bonded warehouse i s o f relatively small scale servicing the need of a single factory. CBW on the other hand will be o f a very large scale supplying inputs to numerous factories. Hence, CBW will be able to reap the economies o f scale that are unavailable to individual bonded warehouses. The large size o f the order o f CBW should enable it to negotiate a lower bulkprice and perhaps lower unit shippingcost. It would also reduce unit administrative and transaction costs associated with imports such as port costs. Part o f the reduced costs o f CBW could be passed on to the domestic exporters. They could also save on interest charges as the local LCs required to purchase inputsfrom C B W would be o f shorter duration. These potential cost savings should improve the competitive strength o f domestic RMGproducers. 7.5 1 However, the operation of CBW byprivate entrepreneurs would also imply certain costs. Large inventories of inputs must be held if the system i s to function efficiently. This obviously has financing cost implications. The entrepreneurs must be remuneratedwith sufficiently highprofit margin for them to continue in the business. These will add to the cost o f the inputs. Double truclung and clearance requirements could also raise the cost. The RMG producers will buy from the CBW only if the cost o f procuringinputs from the CBW i s less than the cost o f procuring the same directly from overseas through their own bonded warehouses discounting for the lead time factor. Ifthis is not the case the C B W cannot be sustained. Indeed, ifsuch an outcome i s foreseen by the entrepreneurs they will not invest inthe CBW business notwithstanding any permission given by the Government; and if they have already incorrectly investedthey will soon go out of business. No public fund should be committed insetting up a CBW. 35 National Co-ordination Council (2004) has suggestedsucha scheme for CBW. 34 7.52 Local export-oriented PTS could also benefit from the CBW. They can procure their raw materials such as yam, grey fabric, dyes and chemicals etc. from CBW. They would also get the raw materials bulk-purchased by CBW at a reduced cost, and could service any orders from local RMG firms at a reducedtime and thereby saving on the leadtime o f RMGexporters. 7.53 CBW can significantly reduce uncertainties of importing inputsfrom overseas.There could be delays in shipment, shipping time, transhipment or port clearance when inputs are imported from overseas. Such delays impose significant costs on exporters as they get less time for production and may be required to ship finished products by air at multiple cost. The delay could also result inthe cancellation of the order. Profit margin i s eroded due to such occurrences. 7.54 Some of theproblems with sourcing arises due to the high protection given to the textile sector. Untilthe beginning ofthis year import of certain categories oftextiles was prohibitedoutright. Although the prohibition has been lifted, the textile sector enjoys fairly highprotection. The effective import duty on many textile products i s as highas 47 per cent. Ifthe protectivebarriers were to be reduced, some o f these problems will fade away. For example, ifimport duties on fabric and yam were abolished, there will be no necessity o f bonded warehouses, either individual or central.36Just as some business houses stock cotton (that i s imported duty free) inorder to sell directly to spinners, so there will be business houses that would stock yam, fabric and other raw materials and intermediate inputs for selling to weavers, processors and RMG manufacturers. The presence o f a tariff wall and the duty-free facility for the exporters prevent such business houses from stocking yam, fabric or other inputs for textile and RMG manufacturers. Hence, elimination o f import duties will have the same effect inreducing lead time as the establishment o f central bonded warehouses in a tariff-ridden economy.37To the extent the immediate suspension o f import duties on textile products i s not feasible for revenue or other reasons, the case for central bonded warehouse has merit. It should be given serious consideration as an instrument o f raising export competitiveness o f RMGsector o fBangladeshinthe post-MFA era. 7.55 The textile manufacturers are predictably not convinced about the merit of CBW. They are staunchly opposed to government giving permissionto set up CBW. Intheir opinion, the PTS as well as RMG industries will be destroyed (sic) if permission were given for the establishment o f CBW as this would flood the market with imported fabric through both formal and informal routes. Their concerns are twofold: (1) CBW will reduce the prices they receive for their yarn and fabric, and (2) it will reduce the size o f the domestic market o f the primary textile sector. If (1) above has any merit then currently domestic yam and fabric must be selling at higher than the international market price (import price) o f these inputs. This is known to be true. Table 12 shows that not only the prices paid by RMG manufacturers for domestic yarn i s 10-15 per cent higher than the prices o f Indian and Pakistani yam, but also about 10 per cent higher than the prices paid by purchasers in the local market. BGMEABKMEA reports that domestic yam manufacturers charge.prices that are often 20% higher than the international prices. The higher prices were until now sustained by generous cash incentives given to fabric manufacturers for using local yam, which constitutes up to 75 per cent o f the value o f grey fabric. Because o f the budgetary constraints and abuse o f the system the Government gradually reduced cash incentives from 30 per cent o f the value o f the fabric to 5 per cent. Inline with yarn prices, the prices o f domestic fabric i s also considerably higher than that o f the imported fabric by a considerable margin (see Table 12). 36 The elimination of tariffs could also see outsourcing of CMT activities to households on a large scale by the garment exporters who would not then needto maintain large manufacturing facilities. This could help themto reduce costs. However, with suchoutsourcing, it would be nearly impossible to monitor social compliance. 37 The converse also holds; if all textile imports were subjected to the same tariffs regardless of the user or use, there would not be any necessity of bonded warehouses. However, in this case the RMG producers will be pricedout ofthe world market. 35 7.56 Since there are no restrictions on imports by RMG manufacturers, one may wonder why they should pay higher prices for domestic yarn or fabric. There are at least three reasons for which the garment manufacturers pay the higher prices for domestic fabric. The RMG manufacturers freely admit that they are quite willing to pay 5-8 per cent higher for domestic fabric because o f the saving they make on the total costs o f importing fabric from overseas. The saving i s due to the reduction intransport costs, port and custom clearance costs and hassles and reduction in lead time. The other reason that permits local fabric producers to charge a higher price i s the European rules o f origin that need to be complied with inorder to avail duty-free access for RMG. Since duty-free imports are only a small fraction o f the total European imports, European domestic prices o f garments are not likely to be influenced by these preferential imports. Consequently, the manufacturers o f RMG who can access duty-free privilege can charge prices higher than the competitors' prices by the margin o f the average tariffs. However, the EU rules o f origin require that the garments from LDCs including Bangladesh must be made from domestically manufactured fabric in order to avail the duty-free access. Most o f the knit garments exported by Bangladesh are made from domestic fabric. As such knit garments enjoy duty-free access into European Union. But only a small fraction o f the woven garment exports from Bangladesh are made from domestic fabric due to the shortage o f domestic supply and hence only this fraction o f the total exports can access duty-free privilege granted to L D C S . The domestic fabric manufacturers thus enjoy a ~ ~ monopoly position with respect to duty-free access into the EU market. This permits them to raise the price o f the fabric to mop up part or all o f the margin on garment exports on account o f the duty-free access. 7.57 The third reason is the reduction in lead time madepossible by the availability of fabric from domestic sources. An order with quick shipment requirement could be serviced only if fabric and other accessories could be sourced domestically. If the profit margin on such quick orders i s adequate, the RMGmanufacturers would be willing to pay a premium on domestic fabric prices. 7.58 The establishment of CBW does not adversely impact on thefirst two reasons for the higher price of fabric. Import costs are probably reduced, if at all, only marginally due to CBW; indeed according to the BTMA presentation to the National Co-ordination Council on RMG Sector o f Bangladesh such costs will rise because o f double clearance and double truckmg requirements. Since C B W will stock only imported fabric, any garments made from fabric purchased from CBW will obviously not qualify for duty-free access to EU. Hence, domestic producers will not pay a premium price for C B W fabric. The latter cannot compete with domestic fabric inthe market for duty-free export. The monopoly position, and therefore, the premium price o f domestic fabric will not be affected by the operation o f CBW. 7.59 However, the establishment of CBW will make importedfabric available as quickly as, in some cases more quickly than, domesticfabric. The RMG producers who are procuring domestic fabric for reducing lead time will be unwilling to pay a premium on this account on the price o f the fabric that i s also available from CBW at the international price. This provides a plausible reason for the opposition o f the textile manufacturers against CBW. The textile manufacturers have an interest to oppose any attempts to reduce the leadtime except through an increase intheir production. As explained earlier, lead time has become an important competitive factor in the post-MFA world and its importance will increase with greater integration and specialization o f the textile market. To remain competitive, and to exploit the full potential for growth, the RMGindustryof Bangladesh mustbe able to reduce the lead time to the level o f its major competitors. The modem textile industry o f Bangladesh owes its growth to the RMG industry. The latter provided the market for the products of PTS. Faster the growth o f the RMGindustry, the larger would be the market o f the domestic textile producers. Any diminution of RMG industry growth will ~~ 38 According to local EU officials, about a quarter o f the woven exports to EUreceive GSP, but BGMEA (2004a) reports that the figure is only 16 per cent. 36 impact on the primary textile industry adversely. The policy o f restricting the growth o f RMG is, therefore, counterproductive; it might bring in higher profit for some o f the current producers by artificially raising the prices o f primary textile products, but it also generates inefficiencies and reduces the market size, and thereby stunts the growth o fthe industryinthe longer term. 7.60 The other concern of PTS is that the setting up of CBW will reduce the demandfor textile products supplied by the domesticPTS industries.RMGproducers will buy these from the CBW instead. This is an admission that the domestic PTS industries are still not competitive internationally. Hence, they must be shielded from the competition o f more efficient foreign producers untilthey grow stronger. This i s the classic infant industry argument which appears to have wide acceptance in the country. The government has provided generous support to the textile sector in order to support their growth. The spinners have been given the benefit o f cash incentives for more than a decade while the fabric manufacturers were protected by either outright prohibition o f imports or very high tariff barriers. The import prohibitions have been phased out at the end o f 2004, butthe highimport specific duties (47.7 per cent in2004-05) are still maintained.39 7.61 Theprotection and support given to PTS have resulted in large increases in investment and production capacity40.Spinning capacity, according to BTMA, has trebledbetween 1994 and 2004 from 1.4 million to 4.2 million spindles. The output o f knit fabric increased from 385 to 990 million meters between 1997-98 and 2003-04 (Table 16). During the same period, woven fabric production increased from 971 to 1430 million meters. Most o f the woven fabric produced domestically i s destined for domestic consumption. Domestic producers supplied less than 15 per cent o f the total requirements o f woven RMG producers (see Table 11). There i s no reliable data about the situation since then, but it i s believed to have not improved much. Domestic producers were, however, much more successful in meetingRMGdemand for knit fabric. By 2001-02, they supplied 65 per cent o fthe demand for knit fabric o fthe knitwear manufacturers. As mentionedearlier, BKMEAclaims that their market share now exceeds 90 per cent. If current trends persist, within a very short period Bangladesh i s likely to become self sufficient inknit fabric production. 7.62 The knitwear manufacturers appear to be satisfied with the quality of knitfabric (Gherzi et a1 2002). The prices are higher than the comparable international prices, but this is offset by the savings on import costs and the higher prices in EU due to the EBA initiative. The knitwear exports have been growing robustly despite the erosion o f the EBA advantage due to the higher prices.41 It was mentioned earlier that CBW will work satisfactorily only if it i s complemented by a flourishing dyeing and finishing industry that can process grey fabric into finished fabric o f the quality that matches those offered by competitors. It i s unlikely that C B W can stock finished fabric o f exactly the colour and design demanded by every single order from overseas buyers. But if the CBW stock grey fabric o f various constructions and there i s sufficient dyeing and finishing capacity, the fabric o f the stipulated construction could be dyed and processed into finished fabric o f the colour and design demanded by the buyer at a short notice thereby reducing the lead time. Since most the fabric demanded are o f some common constructions, it will not be necessary to stock up grey o fnumerous constructions. 39 Purse11(2005). 40 The Government seems convinced that backward llnkage is important which justifies the huge investments made recently in PTS, largely financed through bank loans. Hence, the desire to protect. The point is that it would be much better to protect PTS through direct budgetary subsidies, instead o f dragging down the RMG sector by imposing various policy constraints. 41 This may be overstated. Many exporters now have integrated knit industry that converts yam to garments. Hence, the higher price o f fabric is a transfer betweendifferent branches o f the same fii. 37 Table 16: Domesticproductionofyarn andfabric Yam Fabric (million metres) millionkg All Woven Hosiery & types Knitting 1997198 213 1356 971 385 1998199 229 1424 994 430 1999100 25 1 1599 1091 508 2000101 252 1800 1070 730 2001102 275 2000 1148 852 2003/04 350 2420 1430 990 Source: Gherzi et a1(2002) andBTMA Annual Report 2004. 7.63 The woven section of PTS supplies less than a ffh of the total requirements of woven RMG producers. The rest has to be imported. The problem is then whether the fabric should be imported by individual RMG firms against firm orders (LCs) as i s the current practice, or at least part o f this excess demand should be met through the establishment o f CBW in order to reduce lead time o f at least part o f the exports. To assuage the concern o f the PTS that if C B W are allowed unrestricted import o f woven fabric, the market o f domestic manufacturers may shrink, CBW may be permitted initially to import only a fraction, say 60 per cent, o f the excess demand for each type o f fabric. Such a restriction on the imports of CBW ensures that the market o f domestic woven fabric manufacturers cannot be reduced given their production and projected growth. A similar type o f restriction could be placed on yam, too, so as to protect the market o f the local yam producers. However, such protection should be time-bound to ensure that inefficiencies are not generated inthe sector. 7.64 An of-repeated concern about CBWis that there will be leakages into the domesticmarket that will harm the local primary textile sector. This does not seem to have substance. There are three thousand plus bonded warehouses under the control and supervision o f the individual firms. If leakages from these bonded warehouses have not hurt the growth o f PTS, it i s unclear how leakages from well protected, government andperhaps industry supervised C B W could be hurtful.Inany case, the solution to this problemis tightening of the custom security measures inorder to prevent or minimize leakages. This may be achieved by locating the CBW in the EPZs as suggested by the National Co-ordination Council. The first best option i s always to go at the source o f the problem, rather than controlling its effects through more regulatory measures which could have other undesirable effects. Recommendation: Permit the establishment of CBW in line with the recommendation of the National Coordination Council. Take advantape of external uolicv environment. 7.65 Make the most of preferential market access which gives price advantage. Bangladesh experience shows that preferential access to global markets creates price advantage that translates into more exports. As mentioned earlier, MFA quotas provided sheltered market access to Bangladesh garment exports to U S and EU, prompting the growth and expansion o f an entire industry, and providing job opportunities to hundreds o f thousands o f unslulled labor, mostly women. On top o f it, being a Least developed country (LDC) afforded Bangladesh the benefit o f GSP plus zero tariff access to EU under Everything But A r m s (EBA) scheme, though subject to certain stringent rules o f origin (ROO) requirement. While other developing countries like India and China were restrained both by quotas and non-zero tariffs averaging in excess o f 12%, Bangladesh exports w h c h satisfied ROO requirements entered EUmarkets duty-free, giving them substantial competitive advantage over competitors. 38 7.66 This is illustrated in Table 17. Suppose an apparel product is exported to EU from either a developing or a least developed, or a non-GSP country at $10 per piece. The applicable MFNtariff on the product i s 12.5 per cent. The duty reduction for developing country will be 2.5 per cent but for the least developed country it will be full 12.5 per cent. The non-GSP countries o f course receive no duty reduction. The landed duty-paid price of the product exported from the non-GSP country will be $11.25, that from the developing country $11while that from the least developed country $10. Hence, if all these countries were equally efficient inthe production o f this item, the least developed country will have a 10 per cent price advantage over its developing country competitor and a full 12.5 per cent advantage over a non-GSP competitor. Conversely, even ifthe cost of production o f the least developed country were to be 10 per cent higher than that of the developing country or 12.5 per cent higher than that o f the non-GSP country, its product would still be competitive as the landed duty-paid price o f all these countries would be the same. Ifthe product were priced at $50, the percentage margin per piece remains the same, but the absolute margin becomes larger. Table 17: Illustration of priceadvantage due to GSP scheme Price of MFNtariff GSP for GSP for least Tariff Tariff Tariff inclusive Item developing developed inclusive inclusive price price of least co'untries countries price ofnon- of developing developed GSP countries countries countries $10 12.5 % 2.5% 12.5% 11.25 $11 $10 $50 12.5 % 2.5% 12.5% 56.25 $55 $50 7.67 Thispreferential margin greatly helped to increasethe export of knitwear to EU at a rapid rate. Indeed, more than 95 per cent o f the increase in total knitwear exports o f Bangladesh between 1997-98 and 2002-03 was accounted for by the increased exports to EU. Total knitwear exports were only about one-fifth o f the total exports o f woven garments in 1993-94, but by 2003-04 this ratio rose to more than three-fifths. By volume, knitwear exports now exceed woven exports. EU absorbs nearly three-fourths o f the knitwear exports o f Bangladesh. 7.68 Textile and apparel imports to the North American market (US and Canada) were not subject to GSPpreferences. Consequently, Bangladesh apparel exports to U S faced average tariffs o f about 14%. In2003, Canadaofferedzero tariffaccessandrelaxedROOfor LDCs.As aresult, Bangladesh's exports of apparel went from US$90 million in 2002 to US$342 million in 2004 - a fourfold increase o f exports injusttwoyears. Thistrendis continuing. 7.69 Gaining a foothold and expanding market share in the lucrative US apparel market is every garment exportersdream. Not only does the US not offer duty-free access to LDC exports, its tariffs are generally skewed against export products from developing countries, as a consequence o f which Bangladesh ends up paying more tariff revenues on its apparel exports o f about $2 billion in 2003 than French exporters paid on their exports o f US$26 billion. Under the Caribbean Basin Initiative (CBI) and African Growth and Opportunity Act (AGOA), the U S offered duty-free access to a number o f Caribbean countries and Afhcan LDCs. These concessions put Bangladesh exporters at a price disadvantage and they have been lobbying hard with U S Congress to give them duty-free access under a proposed bipartisanbill called Tariff ReliefAssistance for Developing Economies (TRADE) Act 2005. 7.70 What would be the implications of such a concession to Bangladesh? Using Computational General Equilibrium techniques and the most recent data available42the effects o f elimination o f U S ~ 42 The last versionof GTAF', GTAF' 6.4 ( www.gtap.org)),withthe base year 2001. 39 tariffs on imports of textiles and apparel from Bangladesh has been estimated. The reduction o f tariffs on the U S market shouldincrease the volume o f Bangladesh's exports to US. Table 18 shows the percentage change in Bangladesh exports toward European Union and United States occurring over a period o f 3-4 years. The elimination of tariffs leads to a 90 percent increase in the apparel export to US, and an 82 percent increase in the textiles export to US. The significant increase in exports toward U S markets i s possible by an increase inproduction (as a consequence, the demand for labor inthe Bangladesh's apparel industrywould increase by one third) and by shiftingexports from other markets to the US market (there would be a 15 percent decrease in export o f apparel to EU and a 17 percent decrease inexport o f textiles to EU). The increase in demand for textiles and apparel increases the demand for labor (34 percent increase inthe apparel industryand 7 percent increase inthe textile industry). Table 18: ChangesinBangladeshexports Category EU25 USA Textiles -17 82 Apparel -15 90 Source: World Dank, DECRG Trade Table 19: Welfare changesdue to elimination oftariffsfor Bangladesh`s Exportsoftextiles andapparelto US CountryIRegion Welfare change US$ mil EU25 -36 Chma -27 Bangladesh 476 India 20 Sri Lanka -3 United States -210 Mexico -16 Global change - 205 Source: World Dank, DECRGTrade 7.71 Table 19 shows the welfare effects o f tariff elimination for Bangladesh's exports of textiles and apparel to US. The increase in competitiveness o f Bangladesh exports o f textiles and apparel lead to almost a half a billion U S dollars increase inwelfare for Bangladesh. The loss o f tariff revenue seems to be the main reason of welfare loss for US, the increase in competition inthe apparel market havingjust a marginal effect on the domestic apparel producers. The increase in U S exports i s compensated by a 15 percent decrease in exports towards EUwhich may result ina marginal increase inthe price o f apparel in EUand a welfare loss o fUS$ 36 million.43 The welfare losses o f China andMexico (of US$ 27 million, respectively US$ 16 million) may be explained by the increase in competition in the U S market for apparel and the welfare gain o f India due to the increase in the export o f textiles from India to Banglade~h.~~The elimination o f tariffs results in a global welfare gain o f US$ 205 million as well as large welfare gains for Bangladesh. 43 Bangladesh exported US$ 3.8 billion apparel to EU25 in 2003, roughly 4 percent o f EU's imports (UN COMTRADE). 44 The welfare changes for China and India are influenced by their double role. They are suppliers o f textiles to Bangladesh, thus benefiting from the increase in the demand for textiles due to an increase in the apparel production inBangladesh. On the other hand, they are direct competitors with Bangladesh on the US market. 40 7.72 The decomposition o f welfare changes in variations due to the allocative efficiency or terms o f trade allow for a more precise analysis o f welfare changes (see Table 20). Inthe case o f Bangladesh, the most significant change in welfare i s due to allocative efficiency (US$ 281 million), followed closely by welfare changes due to changes in terms o f trade (US$ 195 million).45 Almost three quarters o f the welfare changes for U S are due to terms o f trade. The elimination o f U S tariffs allows for an increase in the price and volume ofBangladesh apparel exports to US. Table 20: Decompositionof Welfare Changes Countries Allocative Efficiency Terms o f Trade Total Welfare Gains EU25 2 -38 -36 C h n a -4 -24 -27 Bangladesh 281 195 476 India 8 12 20 Sri Lanka -1 -2 -3 United States -56 -154 -210 Mexico -6 -10 -16 Global Welfare 208 -3 205 The table shows just the countries with the biggest variation in welfare. Global welfare cumulate welfare changes for all the countries. 7.73 Although the above discussions have focused on staying competitive and seizing preferential access into the major export markets for RMG (US and EU), this i s not to under-estimate the potential for exports to hitherto untapped markets. Japan, which i s the largest importer o f RMG., after U S and EU, i s yet to be explored, but holds enormous potential. Bangladesh needs to penetrate this market, but faces stiff competition from China. Vigorous marketing and promotion dnves are needed to enter the Japanese, East Asian and other growing markets o f middle income countries, not to mention Australia, New Zealand, and Norway, all o f which have given zero-tariff access to Bangladesh exports. Even Russia and the Middle East hold significantprospects for future exports. Recommendation. Preferential access is time bound and comes with conditions. Bangladesh should seize the opportunities offered to it as an LDC but use the space to preparefor increased competition when it graduates out of the LDC league. The industry should be looking out for other markets beyond US and EL? 7.74 Why not take advantage of Regional Cumulation. An essential requirement for obtaining GSP facilities i s that the export products o f these beneficiary countries must satisfy the applicable rules o f origin. Each developed country (or custom territory) has its own GSP scheme to provide preferential treatment to export products o f developing countries and a set o f rules o f origin to go with it. The discussion on the rules o f origin in Bangladesh usually centers around the EUrules o f origin applying to GSP (EBA) benefits. The reason for this exclusive focus on EUrules o f origin i s that Bangladeshiexports are prevented from fully utilizing GSP facilities because o f ROOrelated problems only in EU. Currently only about one-half of the RMG exports to EU can access GSP facilities. The rest is traded on an MFN basis. Such a ROOrelated problem does not occur in U S A since it does not give any GSP benefits to textile and leather products, and other products such as shrimp easily meet the U S preferential ROO.The 45 Allocative efficiency results from higher Bangladesh exports in response to increased US demand. Terms o f trade effect refer to the rise in the relative price o f exports over imports and the consequent welfare implications. 41 ROOi s also not a problem in Canada and Australia as these are now sufficiently liberal to permit all exports to obtain GSP benefits. There is little export o f RMG products to Japan or Norway and hence, ROOo fthese countries have not emerged as an issue at this moment. 7.75 According to the current EUrules o f origin, exported items o f apparel mustbe `manufacture from yarn', implyingthat ordinary apparel items must be made from domestically produced fabric in order to obtain originating status. This requirement turns out to be fairly stringent. Most o f the exported woven garments to EUare made from imported fabric and consequently do not satisfy the rules o f origin. 7.76 The RMGproducers have lobbied with EUfor the last several yearsfor a relaxation of the aforementioned rules of origin such that garments manufactured from imported fabric could also qualify for GSP benefits, but primary textile producers have opposed any relaxation and the government has also not supported their demand. So far EUhas not relaxedthe rules o f origin.46 However, there exists a small window o f opportunity to circumvent the rules o f origin in the form o f regional cumulation, which permits derogation from the provisions o f Article 67 that define originating rules.47 A beneficiary country, which i s part o f a regional trade area such as SAARC may claim, for GSP purposes, fabric purchased from regional member countries as their own when claiming GSP if the value added there i s greater than the highest customs value o f the fabric used originating in any one o f the other countries o f the regional group and the working or processing carried out there exceeds certain minimumprocessing req~irernents.~~However, when these conditions are not satisfied, the products shall have the origin o f the country o f the regional group which accounts for the highest customs value o f originating products coming from other countries o fthe regional group. 7.77 The cumulation rules clearly indicate that availing thisfacility depends on the cost structure of RMG production and the number of sources of imported inputs used in production. If fabric is imported from a single regional country (as is mostly the case) and all other inputs going into RMG production are sourced domestically, the domestic value addition must exceed 50 per cent o f the ex works price o f the garments to qualify as originating domestic product under the cumulation derogation. This i s unlikely to happen as fabric value i s typically over 60% o f costs. Therefore, Bangladesh needs to request for further derogation from GSP rules o f origin, as was done and granted to Cambodia, Laos, and Nepal. That would allow Bangladesh zero-tariff access to EU as an LDC, while manufacturing operation takes place from there without any value-added condition. 7.78 The current value addition requirement turns out to beprohibitivefor most of the woven RMG manufactures. A number o f studies including Dr.Martelli Associates (1999), Bhattacharya and Rahman (2000) and Gherzi et aZ(2002) show that the value addition inthe production o f the most common items falls short o f the minimum required to qualify for EU GSP. In other words, without a substantial relaxation o f the value addition requirement, most o f the RMG exports to EU will not qualify for GSP even ifregionalfabric is used. Only the highvalue and fashion items, which account for a small part o f the total RMG production could benefit from the existing cumulation rules. Notwithstanding this fact, textile manufacturers have strongly opposed cumulation because o f their apprehension that they would lose market share to Indiantextile producers ifcumulation i s allowed. ~ 46 The inability of a large part o f the LDC manufactured exports to cash on EBA benefits has finally persuaded EU to reconsider the rules of origin. It is now contemplating changes in the rules to a simple value addition criterion such that more exports from LDCs qualify for EBAbenefits. 47 Actually there is another provision that could be used to access EBA. The govement can request for a derogation o fthe rules as an LDC. Such derogationhas been allowed for Cambodia, Laos and Nepal. Not so for Bangladesh as no such request was made inview o f stiff opposition from PTS. 48 See European Council Regulation (EC) No. 2501/2001. 42 7.79 As stated earlier, the local woven textile sub-sector supplies about 15 percent o f the total demand of the woven RMG sector. This sub-sector o f PTS has not been able to increase its share o f the RMG requirement during the last several years since RMG demand has increased at a faster pace than the increase in their production capacity. Unless locally produced woven fabric i s currently beingused only to produce high value fashion items where value addition exceeds 50 per cent and all o f it i s exported to EU, SAARC cumulation can have hardly any effect on the domestic producers. The privilegedposition, and consequently the price premium, enjoyed by the locally produced fabric due to the rules o f origin for EBAwill not be affectedmuchby cumulation. Hence, it appears imprudentto pursue apolicy that holds back RMGproducers from taking advantage o fregional cumulation. Recommendation. Regional cumulation, in whatever shape orform, offers incentivesfor sourcing inputs from the region, which, in and of itself; could be a way for reducing lead time and enhancing competitiveness. why pass up the opportunity? Government should lobby hard with EU to reduce value added requirement along with liberal regional cumulation option. 7.80 An FTA with India could spur RMG exports to the vast Indian market.49 Access to the vast Indian market for readymade apparel without hindrance from tariff and non-tariff barriers could add a significant dimension to India-Bangladesh trade by opening a potential channel for addressing the bilateral trade gap that often i s an eyesore in Indo-Bangladesh relations. An FTA that gives unilateral duty-free access to RMG exports (among others) from Bangladesh, with staggered reduction o f duties for Indian imports, i s an attractive option from the Bangladesh perspective. Whether such an arrangement can be pulled off inthe near term i s a political economy question. Table 21: India andBangladesh: Total Garment Exports 1997-2004 India Bangladesh Fiscal Knitted Not Total Knitted Notknitted Total knitted Year HS61 HS62 H S 6 HS62 1997 1034 2719 3753 763 2238 3001 1998 1023 2855 3878 940 2843 3783 1999 1258 3107 4365 1035 2985 4020 2000 1588 3177 4765 1270 3083 4353 2001 1769 3790 5559 1496 3364 4860 2002 1864 3143 5007 1459 3125 4584 2003 2387 3352 5739 1652 3255 4907 2004 2701 3541 6242 1859 3116 4975 Sources: India: Departmentof Commerce, Export ImportData Bank. BangladeshBank and Export Promotion Bureau. Exports for BangladeshFY04 are up to May 2004 only. The Indian fiscal year i s fromApril 1to March 30. The Bangladeshfiscal year i s fromJuly 1to June 30 7.81 India and Bangladesh are two of the world's leading exporters of readymadegarments (Table 21). The RMG industries in both countries are very large, low cost, internationally competitive and economically efficient export industries. Therefore, the likely economic consequences o f their inclusion ina free trade agreement -whether bilateral or under SAFTA-- are of special interest. This is especially the case in Bangladesh, where many Bangladeshis hope that RMG exports might help reduce Bangladesh's large bilateral trade deficit with India. 49 Discussionon this strategic optiondraws from Gany Purse11(2005), Free trade betweenIndia and Bangladesh: A Casestudy of the ready madegarment industry, World Bank, mimeo. 43 7.82 Butthe economic effects o fopening upRMGtrade between the two countries inthe context o f an FTA are complicatedby: 0 India's prohibitively highspecific duties on most garments 0 Bangladesh's very highRMGand textile tariffs 0 Bangladesh's ambiguous policies on domestic sales by exporting RMG firms 7.83 Tariffs. Before the final removal o f the textile and clothing import ban, with the support o f the Ministryof Textiles, India's T&C industry lobbied the government to impose specific duties on a large number o f textile fabrics and garments. This was done in 2000. In2002/03, the proportions o f H S 6-digit tariff lines subject to specific duties were: cotton fabrics, 49%; man-made filament fabrics 88%; man- made staple fibre fabrics 69%; special woven fabrics (including tyre cord fabrics) 51%; knitted apparel 30%; apparel, not knitted 62%. The tariffs are compound i.e. the higher o f an amount calculated usingan ad valorem rate or the specific amount. The specific component i s in Rupees per square metre or per lulo in the case o f fabrics, and per item (e.g. per shirt) inthe case of garments. For the products subject to specific duties, the objective and effect i s to target and keep out imports o f specifications and qualities that sell for l o w prices for which there i s the largest demand in the Indian domestic market, and o f which other developing countries are the most competitive foreign suppliers. The specific tariffs were initially set at very high levels during FY 2001: they were reduced somewhat in FY 2002 and have remained at that level since.50 Table22: India. Mens' or boys'woven cotton shirts: MFNtariffs and SAPTA preferential tariffs on imports from Bangladesh. Comparisons of ad valorem and ad valorem equivalents of specific tariffs, 1999/2000-2005/06 Ad valorem rates Specific tariff ad valorem equivalents $3 shirt $4 shlrt $5 shirt MFN Pref MFN Pref MFN Pref MFN Pref 1999100 45.7 45.7 45.7 45.7 45.7 45.7 45.7 45.7 2000101 44 24 106.5 55.2 80.8 42.4 65.5 34.7 2001102 40.8 22.6 66.3 35.3 50.8 27.6 41.5 22.9 2002103 35.8 20 65.4 34.9 50.1 27.2 41 22.6 2003104 35.8 20 68.6 36.5 52.6 28.4 42.9 23.6 2004105 20 10 63.1 31.6 47.3 23.7 37.9 18.9 2005106 15 7.5 63.1 31.6 47.3 23.7 37.9 18.9 Notes: There were no SAPTA preferences for garments in 199912000andbefore, andno specific duties.From 2000101the Bangladesh SAPTA preference has been 50%, butthat only applies to the Indian"Basic'' Customs duty. Consequently, until 2004105 the preferential tariffs were more than 50% o f the MFNtariffs owing to the application of other protective import duties The last o f these other duties (the Special Additional Duty or Sadd) was removed inJanuary 2004. A small "education cess" (2% ofallimport duties) introducedin2004hasnotbeenallowedfor. 7.84 Under SAPTA, preferences for garments were introduced in FYOl-but for "least developed" SAPTA countries only. Subject to meeting rules o f origin, the preference for Bangladesh and the other SAPTA LDCs (Nepal, Bhutan andMaldives) onmost garments is 50%, so the basic Customs dutywould be the greater o f an amount calculated by applying half the ad valorem rate to the c i f price, and half the 50 The FY 2003/04 Customs tariffschedule extended the HS classification systemfrom 6 to 8 digits, and indoing this the specific duty for any given6-digit product was also usedfor its 8 digit subproducts. 44 specific amount5'. There are no garment tariff preferences for the other "developed" SAPTA members i.e. Palustan and Sri Lanka. However, under the bilateral agreement with Sri Lanka (ILFTA) the preference for garment imports from Sri Lanka i s 75%, and garment imports from Nepal (like imports o f nearly all other products) are duty free. In order to qualify for the SAPTA preferences, garments imported from Bangladesh would have to satisfy the SAPTA origm rules. For the SAPTA LDCs the principal requirement i s that the cif value o f non-SAPTA imported inputs included in the exported product should not exceed 70% o f the fob price, or put another way, that national value added should be no less than 30% o f the fob price. 7.85 Prospectsfor Bangladesh garment exports to India. Three important factors which affect the ability o f Bangladesh garment firms to export to India with the present SAPTA preferences or hypothetically inthe future with a free trade agreement, are: (1) production costs inBangladesh relative to production costs in India; (b) India's tariffs in relation to the Bangladesh cost advantage, if any; and (3) whether and how the Bangladesh firms satisfy rules o f origin. 7.86 Production costs: Bangladesh vs India. Studies o f export firms in major garment exporting countries consistently indicate the Bangladesh has the lowest wages, and that the advantage this should give Bangladesh garment firms over firms in other developing countries (including India) i s not offset by lower labour productivity. Table 23 shows some comparisons taken from a recent (2002) report on the RMGindustryinBangladesh. Labour cost Minutesper Impliedlabor cost Minutesper pair Impliedlabour cost $US $Ushour woven dress shirt $US/shirt ofjeans /pair ofjeans Bangladesh 0.23 61 0.23 62 0.24 India 0.41 63 0.43 65 0.44 Pakistan 0.37 65 0.40 70 0.43 Sri Lanka 0.35 50 0.29 55 0.32 China 0.77 60 0.77 60 0.77 Indonesia 0.41 n.a. n.a. 70 0.48 Vietnam 0.30 65 0.33 65 0.33 Italy 14.71 35 8.59 35 8.59 Notes: Labour costs andproductivity estimates from Gherzi report (2002). Impliedlabour costs calculatedfromthese numbers.The productivitycomparisonsare for the same technologylevels except inthe case ofItaly 7.87 The implied direct labour costs derived from these comparisons suggest that for dress shirts and jeans Bangladesh has a labour cost advantage o f about US 20 cents per unit over India. Generally lower wage levels in Bangladesh also presumably mean that-provided productivity is similar or better- clerical and other overhead labour costs are also lower in Bangladesh than in India, and this would also apply to transport and other expenses if (a big if) these services were equally or more efficient. Even so, relative to typical selling prices, the likely cost differences are minor and could easily be offset, or more than offset, by differences in other costs, or in quality, design, delivery times and other non-price factors. Certainly, they do not provide a decisive cost advantage comparable to the cost advantage that ~~ 51 The preferences did not apply to India's "Special additional duty" (Sadd). In most circumstances the Sadd provided significant extra protection, so the preferential protection ratewas greaterthan impliedbythe apparent preference. For example, in 2002103 the MFN ad valorem protective rate on garments (including the effect of the Sadd) was 35.8% and the preferential protective rate on imports from Bangladeshwas 20%. The Sadd was abolishedinJanuary 2004. 45 Bangladesh and other developing countries have over garment producers in developed countries, as illustrated by the direct labour cost per shirt in Italy ($8.59) versus U S 23 cents per shirt in Bangladesh. Relative to prices, the 20 cents production cost advantage over India i s only 6.6% o f the price o f a shirt selling for $3, and 4% o f the price o f a shirt selling for $5. Relative to margins (fob prices over input costs), the difference i s greater: about 20% o f typical woven shirt margins o f about $ l / ~ h i r t but ,even if ~ ~ there were an FTA this difference still appears to be too small to be a major influence on sourcing decisions. 7.88 Cost advantages and Indian tarvfs. It i s also instructive to compare the apparent labour cost advantage o f garment production in Bangladesh with India's tariffs, especially the specific tariffs. For woven shirts, for which the apparent labour cost advantage i s about 20 cents, in 2004/05 the Indian specific tariff on a shirt imported from Bangladesh was 92 cents, and on a man made fibre shirt it was about $1.3 1.These tariffs far outweigh the Bangladesh labour cost advantage, and unless the Bangladesh shirts had some special style, brand, marketing or other advantage, would appear to preclude imports o f low value shirts, at least on a substantial scale. As discussed previously, most o f the garments which Bangladesh exports to developed countries in large quantities are subject to specific duties in India, and even though Indias SAPTA preferences for Bangladesh cut most of these duties by 50% (and a few by 60%). the preferential duty in most cases is still much too high to allow substantial exports from Bangladesh to India. Table 24: Pricesof some BangladeshexportedRMGsand preferentialtariffs for BangladeshimportsinIndia. Prices& tariffs in$US/garment Ad valorem Export CMT Specific duty price price duty 7.5% Polo shuts 2.53 0.96 0.98 0.19 T - s ~ u ~ s 1.45 0.38 0.52 0.11 Sweaters 3.58 n.a. 0.98 0.27 Jogging suits 9.79 n.a. none 0.73 Nightwear 3.5 n.a. none 0.26 Pyjamas 2.33 n.a. none 0.18 Notes: prices from Gherzi report. Ch 4, p.64. H S classifications are not given, so the Indian tariffs may vary from these e.g. depending o n the predominant material used 7.89 Table 24 shows current Indian 2005/06 tariffs in relation to typical export prices o f a few Bangladesh RMGs.The specific duties on knitted polo shirts and T-shirts, which are two o f Bangladesh's highvolume exports, are higher than typical CMT prices i.e. the prices quoted to cut, make and trimfrom fabrics provided by the buyers, so it i s unlikely in the extreme that an Indian trader would ever source these garments in Bangladesh if comparable prices could be obtained from suppliers in India (and even less likely to source from China or some other MFNsupplier for which the specific tariffs are double the rates applied to Bangladesh e.g. $1.04 for a T-shirt selling for $1.45 fob and with a C M T price o f U S 38 cents). The products just subject to the preferential 7.5% ad valorem would have a better chance o f competing in India, but for low value products even these duties would probably make life difficult for Bangladesh suppliers. For example, if India were to remove the specific duty on Bangladesh T-shirts, the ad valorem duty for Bangladesh suppliers would still be around 30% o f the CMT price. For these reasons, given the highly competitive Indian RMG industry, Bangladesh garment exporters would probably need duty free access to India to have much hope o f winning substantial market shares inhigh volume products. IfIndia abolishes the specific duties, even at the current reduced ad valorem tariffs and 52 Gherzi report Table 4.3 46 with a substantial advantage over MFN suppliers, export prospects would be modest, and if the present specific duties are maintained, Bangladesh suppliers would probably be confined to small, mostly specialised or opportunistic fringes o f the IndianRMGmarket. 7.90 Rules of origin. For the SAPTA LDCs (Bangladesh, Nepal, Bhutan and the Maldives) to obtain tariff preferences in exporting to India, the maximummaterial input content (valued at c i fprices) that can be imported from non-SAPTA countries, i s 70% o f the fob price. In this respect, Bangladesh has an advantage over Palustan and Sri Lanka, for which the maximum imported content i s still 60%. As i s usual inpreferential trading areas, the minimumlocal content requirement can be met by importing materials from other SAPTA members, but subject to the additional constraint that if this i s done, the content requirement increases to 40% o f the fob price. For Bangladesh firms exporting to India, this means that if they use imported materials from India to help satisfy the SAPTA origin rule, they have to buy sufficient o f these materials to raise the regional content (i.e. processing costs and Bangladesh materials, plus the materials imported from India) to at least 40% o fthe fob price. 7.91 As India is in any case supplying a substantial share o f the imported inputs o f Bangladesh's exported RMGs, satisfying the SAPTA rules o f origin i s probably not at present a serious problem for most Bangladesh RMG exports to India. Moreover, backward integration inknitting already allows most Bangladesh knitted garment producers to easily satisfy this requirement. Therefore the application o f the same or similar origin rules in a bilateral India-Bangladesh FTA, or in SAFTA, would constrain Bangladesh exporters to some extent, but not seriously. Nevertheless, if Indian materials that would otherwise not be used, are used to satisfy the origin rules, by definition the Bangladesh exporters to India are disadvantaged relative to a situation without the rules o f origin. Inthis regard, the Bangladeshban on yam imports from India by the land route (in force since 2002) obviously disadvantages Bangladesh RMGfirms (especially knitwear firms, it seems) exporting to the rest o f the world and actually exporting (or interested inexporting) to India. 7.92 In the preceding sections, we have tried to demonstrate that an FTA with India, with both countries maintaining their present tariff and other policies with respect to the rest o f the world, would provide export opportunities in India for Bangladesh RMG products. This was demonstrated using the example o f mens' and boys' cotton shirts, which are among Bangladesh's principal exports to developed countries and which are being exported to India in small volumes over India's specific tariff, but with the assistance o f a 50% SAPTA preference. This i s done on the assumption that initially there would bejust a one-way agreement, in which India would grant duty free access to Bangladesh RMGs, but Bangladesh would for the time being not provide any preferences for RMG imports from India, or stagger duty elimination for a few years. 7.93 However, a number o f cautionary points needto be made inthis regard: 0 RMG industry in both India and Bangladesh are highly competitive. Domestic prices o f exported garments are probably close to or below cif import prices. Therefore, even with the advantages o f tariff exempt access and proximity, Bangladesh garment exporters would have to mainly compete for market niches based on styles, design, marketing, quality etc. For this they would need to work with Indian traders and distributors, and the Indian traders would have to find some clear advantage in sourcing from Bangladesh rather than from Indian suppliers. 0 A number o f RMGs which Bangladesh exports to other countries are not subject to specific duties in India, but until FY 04 none were being exported to India. For most RMGs India's preferential ad valorem tariff for Bangladesh in 2003104 was 20%, and after that it was reduced to 10% in FY 05 and in FY 06 to 7.5%. Unless some Bangladesh exports have 47 developed since FY 04, this suggests that even relatively low ad valorem tariffs are sufficient to deter exports o f these products to India. 0 Sri Lanka i s a major RMG exporter and under ILFTA it has more generous RMG tariff preferences (mostly 75%) inIndia than Bangladesh's SAPTA preferences (mostly 50%, some 60%). In 2005/06 its preferential ad valorem tariff for RMG products was 3.75%. Despite this, its RMGexports to India are practicallynil. 0 Garment industry direct labour costs are much lower in Bangladesh than in India, but these differences are minor relative to typical garment selling prices in world markets, and in selling to India could easily be offset, or more than offset, by differences in other costs (especially fabric and yam costs), quality, design, delivery times and other non-price factors 0 IfBangladesh garment exporters were to develop substantialexports ofsome RMGsto India, part o f the new exports are likely to have been divertedfrom other export markets. Hence, the net increase in exports resulting from the preferential opening o f the Indian market will be less than the increase inexports to India. Finally, it i s useful to take note that Bangladesh has signed on to the framework agreement for Bay o f Bengal Initiative for Multi-sectoral Technical and Economic Cooperation (BIMSTEC) Free Trade Area, whose constituent agreement for trade ingoods i s slated to take effect inJune 2006. This brings into focus a middleincome country like Thailand as a potential market for RMGwith zero-tariff access. Recommendation. As a negotiating strategy, Bangladesh would be well advised to have a short negative list for SAFTA or FTA, as long as RMG is not on India's negative list. Penetration of the Indian RMG market via preferential arrangements remains a viable strategic option, as also are potential openings throughBIMSTECfree trade area. Compliance with social and labor standards 7.94 As Bangladesh emerges as a major player in the global RMG export market, the pressure i s mounting for compliance with social and labor standards. Failure to conform to internationalnorms could undermine competitiveness. The European Union and the United States, which together account for more than 70 percent o f Bangladesh's total export market, insist on strict compliance with social and labour- related standards, as well as ecological norms, as a pre-condition for doing business. In the face o f international competition inthe post-MFA era, Bangladeshi manufacturers must be able to respond to and harmonize their views with buyers inorder to address the challenge. 7.95 Evidence suggests that hitherto inadequate attention has been given by the industryto the various aspects o f social accountability. It would now be opportune for Bangladeshi firms to conform to SA8000 standards - the international standard for social accountability in the workplace. SA8000 provisions focus attention on the personal safety and health of all employees recognizing that contented workers are motivated to raise their productivity and work quality. Though designed to cover welfare and comfort o f employees, SA8000 has now been adopted, with some variations, as a benchmark for sourcing decisions bybuyers53.Aspects coveredby these standards include fair compensation, compliancewith labor laws - e.g. minimumworking age, work hours, good personnel practices, environmental protection, hygiene and sanitation, etc. They also relate to labor safeguards to ensure health and comfort o f employees in the workplace: e.g. work posture, temperature, humidity and air circulation, illumination, fire prevention and firefighting, etc. 53 The problem is that, unless adaptedto country circumstances, SA8000 couldbecome a ploy to protect highcost fmindevelopedcountrymarkets. 48 7.96 Itis advisable, therefore, that RMG firms recognize these requirements asnecessary for the health and safety o f employees and undertake that the standards set out in SA8000 are implemented and maintained. This is important not only from the point o f view o f enhancing competitiveness and sustaining repeat orders, but i s also critical for keeping workers motivated to improve productivity and product quality. 7.97 BGMEA, in cooperation with EO, has been working towards meeting social compliance requirements in its member factories. Besides LO, development partners, like UNDP, GTZ, and EU, are workmg with manufacturing enterprises, civil society, buying houses, labor organizations and the Government to effectively raise the capacity o f compliance services in Bangladesh. The Ministry o f Commerce has taken the initiative to form a "National Forum on Social Compliance in the Textile and Garment Industry", along with issue-based "Task Forces on Social Compliance", and a "Compliance Monitoring Cell". All this augurs well for the sector under the post-MFA regime. Recommendation: To remain competitive and win orders, RA4G Jirms must conform to international standards of social compliancefor their industry. Government-industw vartnershiv to develop forward linkaaes 7.98 Most assessments of the Bangladesh RMG industryfind serious shortcomings with regard to its international market orientation - the ability to gear up its marketing campaign in response to changing demand. Long years of exclusive reliance on foreign buying houses or their agents have contributed to this outcome. The post-MFA trading regime demands savvy merchandising and marketing o f products in order to penetrate and maintain market share in export markets. Focus on such forward linkages i s a must. The following approaches highlight some o f the key strategies for promoting forward market integration: 7.99 Product and market diversification: One o f the major limitations to apparel export growth is the lack o f diversification inproducts and markets. Besides the export concentration that has been discussed earlier, there i s also a problem o f product concentration - a small range o f products (shirts, trousers, T- shirts, sweaters, jackets) make up 60% o f RMG export. Vigorous marketing and promotional drives would be neededto break into markets for other RMGproducts that Bangladeshi producers are capable o f producing competitively. That brings up the critical role o f R&D and market research to expand the product mix and public support inthis area might be needed. A comprehensive research centre built on public-private partnership should be able to gather and disseminate information effectively to local manufacturers on the latest developments in products and markets, including information on fabric developments, blends, colours, patterns, latest fashion trends and design forecasting, as well as providing customer service to foreign buyers purchasing from Bangladesh. 7.100 Image building and brand development: Image building and branding can achieve a high level o f value addition. As part o f image building, brand development can also enhance Bangladesh's reputation as a quality supplier o f apparel. As branding i s an expensive investment, incentives and opportunities should be provided to exporters, such as through a `Brand Fund' and encouraging foreign collaboration to launch collective brand names through corporate marketing companies. Investments in skills, design and advertising will have to be made in order to make the yet unknown "Made in Bangladesh" label become known inthe global marketplace. For this, an active promotion campaign will be required by the industry, Government and Export Promotion Bureau. Local and international exhibitions should be held with aggressive efforts to attract foreign buyers. Bangladeshembassies abroad should appoint dynamic business oriented commercial officers whose mainjob will be to promote export products through direct contact with potential buyers. Without being pro-active, it i s impossible to build direct relationships with buyers. 49 7.101 At home, an exporters' performance rating database could also be developed, based on factors such as product quality, timely delivery and financial performance. This facility should be set up in collaborationwith an accredited international company so as to ensure objectivity and credibility. 7.102 FDI or Joint ventures: Bangladesh has a limited number o fjoint venture RMG industries; but these can be very effective inbringinginmanagerial and marketing expertise. Continuous efforts should be made by the Export Promotion Bureau and Board o f Investment to ensure a favourable investment climate to allow for more FDI or joint ventures. The positive spillover effects o f FDI has already been noted. 7.103 Specialized units: The RMG sector constitutes a long value chain, ranging from spinning, dyeing, weaving, printing,stitching and knitting. Eacho f these processes involves independent units with specialized labour and machinery. Small and medium enterprises should be encouraged to set up these specialized units, with incentives such as provision o f long-term financing at subsidized rates and allowing duty-free imports o f the machineryrequired. This calls for supportive public policy. 7.104 Human resources development: Middle-level management i s a bottleneck in the industry and there i s a critical need for skill enhancement. Focus should be placed on `training o f trainers' to ensure continual quality training and s h l l development. 7.105 Itis relevant to note that FDI, specialized units and HR development are all important topics by themselves. But the report does not elaborate on these topics as they would require very detailed treatment. Recommendation: Bangladesh RMG industry must gear up to meet the many shortcomings that create a distance between itself and its markets/consumers. Building strong buyer-supplier relationships and focusing on image building (quality, skills, branding, etc.) are two areas most in need of attention throughpublic-privatepartnership. Emulovment Optionsfor disulacedEarment workers 7.105 Post-MFA threat ofjob loss. What needs to be done. RMGhas been the driver o fjob growth in Bangladesh. As such, there i s justifiable concern regarding the employment effects o f quota elimination. Though nothing i s certain, there i s general agreement that there will be some disruption o f employment with consequent labor market adjustment. Transitional adjustments within the RMG workforce i s not unknown. An internal BGMEA survey recently found that o f the 4300 listed firms intheir database, some 1300 were either closed or inoperative prior to the MFA phase out. That would put an estimated 700,000 workers injeopardy. However, in light o f expanding demand, the remaining 3000 BGMEA firms have expanded jobs and exports, and, according to BGMEA, as many as 84 new firms were established in FY05. In all likelihood, most o f the laid off workforce from the 1300 closed firms will have found alternative RMG employment or, as discussed later, moved to home based work. Consequently, most analysts find little evidence o f floating unemployed garment workers. Whether or not this situation will change depends on how the industrycopes with the competitiveness challenges o f the future. 7.106 Earlier, we estimated the RMG workforce to be around 2.0 million o f whom some 80% were females. Another 0.8 million are engaged in accessory and linkage industries. Overall, some 10 million people depend for their livelihood on the RMG sector. That i s a substantial proportion o f the nation's population whose livelihoods might be at risk if post-MFA developments have adverse implications for the sector. Using CGE global models, the World Bank (2005b) estimates potential job losses to the tune of 17% o f RMG employment (about 0.3 million) over the next three-four years, assuming no change in productivity or competitiveness. However, if duty-free access to U S A could be obtained under the 50 TRADE Act, this could alone generate 0.6 million jobs. By implementing some o f the strategic options identified for the sector, we have argued that these job losses could be prevented or even reversed. Because of the predominance o f females in the workforce, some negative social outcomes o f massive job losses cannot be ignored and steps needto be taken inadvance to forestall any adverse fallout. 7.107 Labor market adjustment towardshome-based work. Because o fthe potentially disproportionate impact on women of any adverse change, alternative employment opportunities ought to be explored. Surveys reveal that there i s some tendency for retrenched garment workers to move to home-based informal employment. The pressure to achieve competitiveness globally i s likely to push down labor cost further, leading many workers to move towards more informal setting e.g. home-based work54.There i s evidence that there i s more sub-contracting for piece work, and a portion of it is going to home based workers. In 1990, a BIDS survey found that 44% o f factories were manufacturing their own products without sub-contracting, which was reduced to 27% in 2003. Recent reports indicate that majority of the retrenched garments workers have moved to home-based informal work, without any formal protection or support55. However, movement o f factory workers towards home-based work has become a common practice in most garment exporting countries. This has been the trend in Thailand and Sri Lanka, where home-based workers are a substantial proportion o f RMG workforce. India, a major competitor o f Bangladesh RMG in global market, has a long tradition o f home-based workers in garments and textile industries. InIndia, a large number o f home-basedwomen work inthe informal sector, producing fashion products and customized materials for global export. The home-based industry in India have a major research and development (R&D) components, so this sector i s introducing new fashion, design, ethnic textile and embroideries, which i s captivating the global highfashion market. These women have formed cooperatives like SEWA in Gujarat, and are able to conduct collective bargaining agency (CBA) functions with buyers, and get wages and benefits close to formal sector. SEWA was able to establish their own credit system through bank's and marketing information and system. Bangladesh NGOs are famous for micro-credit, which promoted income generation via micro-enterprise for women. However, these women have not been able to form cooperatives like SEWA, in order to access large scale loans, obtain market information, and engage in collective bargaining with buyers. As current trends suggest significant movement of women workers towards informal home-based markets, it has become important to provide them with credit, technical and managerial slulls, and appropriate institutions so that they are able to expand their enterprises and enhance their well being. 7.108 Reverse migration. Though the number o f home-based workers i s still not substantial in Bangladesh as in other garment exporting countries, there are pockets where one could find workers working from home, especially embroideries in knitwear. A study on garment workers who were retrenched after 2001 found that 60% has become home-based workers and doing embroidery and other piece rate work for garments industry. Some o f them have gone back home in Gazipur, Netrokona and Mymensingh, where some RMG industries have also moved. These women either found RMG jobs in their home districts or are doing piece work from home. There has been a trend by RMG owners, especially knitwear, to move factories towards small towns, where skilled workers live. Many o f these moved towards areas with concentration o f tribal population, as they are known to be hardworking, honest and usually do not have much work, but are unwilling to move from their ancestral land. Therefore, some experts feel that one possible adjustment to post-MFA labor market shocks could be a reverse migration by workers who move back to their own localities and work there. 54 Paul-Majumdar P. (2005) "Bangladesh RMG Moving Towards Informal sector". BIDS Annual Journal, Vol. 22. 55 ibid 51 7.109 A recent survey on RMG workers found that about 80% of workers know about MFA phase out and its potential implications on their l i ~ e l i h o o dT~o~conclude,, for potential laid o f f workers, there are . three possible adjustment options - (a) find other jobs or income opportunity in the RMG sector in the cities, (b) return home inrural areas, or (c) migrate abroad. Among women interviewed, 68% would like to stay and find opportunities inthe city, 10% will go back to rural areas, and 12% would try to go abroad as they have been saving for this purpose. Majority o f workers will look for jobs in other industries, especially garments or become self employed by using their sewing slull. Women who are planning to go back to rural areas expect to initiate small enterprises inlivestock, poultry, and cash crop by talung micro- credit from NGOs. Recommendations Facilitating home-based work is an option that will help labor market adjustment should there be adverse impact ofpost-MFA. Skill development and refreshers training: This would be strategic for accommodating laid off workers as well as new entrants inthe labor force. Therefore, skill development andrefreshers training, consistent withmarket demand, shouldbe ensured. .For developing such trainingmodules, market survey for slulls should be undertaken. Labor Market Information (LMI)System: Most workers do not get severance pay or notice when retrenched. Ifgovernment wants to provide a safety net package, it would be difficult to trace the retrenched workers. Therefore, Labour Market Information (LMI) System will be necessary, within which all workers would have IDcardwith registration number and home district address so that they can be . traced ifretrenched. VIII. Conclusions 8.1 The end of M F A presents both challenges and opportunities for the RMG sector of Bangladesh. There are many uncertainties about new market developments inthe coming months. What i s certain i s that there will be heightened competitive pressure with the more efficient producers attempting to capture a larger share of the global market. Signs o f this are already visible. To sustain and enhance Bangladesh's share, all stakeholders including the government, industryand individual firms will have to quickly respond to the emerging situation with the right policies and actions in order to become more competitive. As stated at the outset, the cost o f inaction or policy errors could be severe. Delays to adopt the right policies, or remove the constraints, can only raise the cost o f missed opportunities and reduce the competitive strength of the sector that has made a remarkable contribution to industrial employment, woman empowerment and reduction o f poverty. It would be unfortunate if the growth potential o f the sector i s stifled or the hard earned gains whittled away because o f policy failures. More important, it would be a grave mistake to take any policy andor institutional action that might undermine the RMG sector's export competitiveness, and any existing trade policy distortion that i s harming this sector's competitiveness would needto be removed urgently. 8.2 I n thepost-MFA world, lead time has emerged as a crucial determinant of competitivenessor attractiveness o f the vendor to the buyer in apparel trade. Several new developments inthe retail business have contributed to a sharp decline in the lead time in USA and Europe. Bangladesh woven RMG Suhrawardy G.M., Ghosh, P.S. and Hossain, M. A. (2004), "Post MFA -PotentialDisempowerment of RMG women workers". BangladeshEconomicAssociationConference, December 8-10,2004 52 manufacturers have one o f the longest lead times amongst competing countries. The principal reason for the long lead time i s the unavailability of local woven fabric. There i s little prospect of changing the situation in the short to medium term given the very large investment requirement and an adverse international market situation characterized by over-capacity and low prices. Hence, some innovative scheme needs to be put inplace to shorten the lead time. One such scheme i s the establishment of central bonded warehouses. CBW can reduce the lead time of local manufacturers to the level o f competing countries. However, to reap the full benefits o f CBW, these would have to be complemented by good processing facilities. Appropriate policies to set up CBW with adequate safeguards against leakages, such as that suggested by the National Coordination Council, should be implemented to ensure that the local woven RMGmanufacturers do not lose out to competitors because o f the long leadtime. 8.3 There is a suggestion that the government should activelypursue policies to quickly increase the capacity of the woven PTS such that they can meet the entire demand for woven garments. To make thispossible, the argument goes, the protectivemeasures for the PTS shouldbe maintained. Some caution i s needed inresorting to such a course. The government in any country i s not inthe best position to judge or accurately forecast the success o f a particular business and hence, it will not always correctly choose winners. It i s more likely to bendto lobby pressure. The best course for policy action for a government i s accordingly to take an even handed approach to all industries by removing any constraints to fair market competition. Studies have been completed inmost South Asian countries in order to formulate a strategic vision for the apparel-textile sectors under the post-MFA regime. In these studies, the potential for the apparel sector to lead the thrust for textile growth has been recognized. So i s the case inBangladesh. This brings up the strategic importance of domesticpolicies to ensure a smooth supply of raw rnaterialshputs to the apparel sector at internationalprices. 8.4 Past experience suggests that trade preference given to the LDCs by the developed world can play an important role in promoting their exports. Currently RMG products fiom Bangladesh are not eligible for the duty-free access to the U S market, but such access i s available to the exporters o f African, Caribbean and several other countries. The slow growth o f exports to the USA (relative to EU) i s partly explained by this lack o fpreferential access. Efforts should be made to ensure that the bill to secure duty- fiee access to the U S market for the exports o f several least developed countries including Bangladesh, now under consideration o f the U S lawmakers, i s passed. The support o f international organisations should be courted to achieve this result. 8.5 Most of the woven RMG exports to EU do not qualijjfor duty-free entry as they do not meet the rules of origin. Ifthese rules were relaxed, such as by reducingthe process requirement from two to a single stage or by applying a sufficiently liberal value addition criterion, the woven exports would qualify for the duty-free access. This would give a boost to the sector for more rapid expansion. Currently the EU i s considering adopting a simple value addition criterion that would help the beneficiary countries to better utilize the GSP facilities. The government, together with industry, should mount a campaign to convince EU authorities that the criterion actually adopted is sufficiently liberal to allow the woven manufacturers to avail duty-free access. 8.6 Under the existing rules of origin exports of RMGproducts madefrom importedfabric could access duty-free facility under EBA only through regional cumulation. If the value addition criterion actually adopted i s not sufficiently liberal, the option o f accessing duty-free facility through cumulation should be kept open. Cumulation will not reduce the opportunities for the expansion o f the domestic woven textile industry given that it supplies only a small fraction o f the market demand. Significant expansion of RMG production can only drive up the demandfor domesticfabric providing the needed stimulus to theprimary textilesector. 53 8.7 Comprehensivelyfacing the post-MFA challenge requires actions on thepart of Government, private sector collective bodies and individual firms. Clearly, some of the required actions identified in the report are in the nature of public goods (e.g. textiles protection policy) and hence are best dealt with by government throughpublic policy, including public investment; some are club goods (e.g. promoting social compliance) and are best addressed by the private sector collectivebodies (BGMEA, BKMEA, and BTMA); while some others are private goods (e.g. striving for cost competitiveness) and should be addressed by individual firms. Only then will the overall outcome exceed the sum of these constituent endeavors and what mighthave been a zero-sum game becomes a positive sum game. 54 References Abernathy, F., et. al. 1999. A Stitch in Time. Oxford UniversityPress, New York. Aitken, Brian J., and Ann E. Harrison. 1999. "Do Domestic Firms Benefit fiom Direct Foreign Investment? Evidence from Venezuela." American Economic Review Vol. 89, (3): 605-618. Bangladesh Export Diversification Project (BDXDP). 2001. Report on Impact o f SAARC Cumulation Agreement on BangladeshRMG and Textile Sectors and on the Overall Economy of Bangladesh. Dhaka, April. Bangladesh Garment Manufacturers and Exporters Association. 2004. "Introducing Central Bonded Warehouses inBangladesh: Rationale, Benefits, Operation, Risks and RiskMitigants." Dhaka, April. .2004a. Post-MFA &Bangladesh: TheImperative of Readiness.Dhaka, May 30. Bhattachariya, D, and Rahman, M. 2000. "Experience with Implementation of WTO-ATC and Implications for Bangladesh." CPD Occasional Paper Series, No. 7. Centre for Policy Dialogue, Dhaka, September. Business for Social Responsibility. 2004. The Multi-Fibre Arrangement (MFA) Strategic Sourcing Impact: ThePrivate Sector Perspective. October 2004 Cookson, F.2003. "Facing 2005-Outlook for the Garment Sector." Dhaka, March. CPD. 2003. Phasing Out the Apparel Quota: Addressing Livelihood Concerns in Bangladesh. University PressLimited. Dr. Martelli Associates. 1999. Bangladesh :Textile Study (Phase II),Interim Report. IFC, Washington DC, May. Eaton, Jonathan, Samuel Kortum, and Francis Kramarz. 2004. "Dissecting Trade: Firms, Industries, and Export Destinations.'' American Economic Review. Elbehri,A. 2004. "MFA Quota Removal and Global Textile and Cotton Trade: Estimating Quota Trade Restrictiveness and Quantifying Post-MFA Trade Pattern." Prepared for the 7' Annual Conference on Global Economic Analysis. Washington DC. European Union. 2002. The European Community's Rules of Origin for the Generalised System of Preferences: A Guidefor Traders.European Commission, July. Post-MFA Development Strategy and Technical Assistance for the RMG Sector: Volume-1." Dhaka, Gherzi Textile Organisation and Project Promotion and Management Associates. 2002. "Initial Report on August. , 2002a. Executive Summary of the Initial Report on Post-MFA Development Strategy and Technical Assistance for the RMG Sector, report submitted to Ministry of Commerce, Government o f Bangladesh. 55 Islam. 2001. The Textile and ClothingIndustry of Bangladesh in a Changing World Economy. University Press Limited. Javorcik, Beata. 2004. "Does Foreign Direct Investment Increasethe Productivity o f Domestic Firms? In Search of Spillovers through BackwardLinkages." American Economic Review Vol. 94 (3). Kathuria, S., Will Martin, and Anjali Bhardwaj. 2001. Implications for South Asian Countries of Abolishing the MultijibreArrangement. World Bank. Kathuria, S., and A. Bhardwaj. 1998. Export Quotas and Policy Constraints in the Indian Textile and Garment Industries. World Bank Policy ResearchWorking Paperno. 2012. WashingtonDC. Kee, Hiau Looi (forthcoming). "Firm Performance in the Services Sector." Malaysia: Firm Competitiveness,Investment Climate, and Growth.World Bank. Kee, Hiau Looi, 2005. "Foreign Ownership and Firm Productivity in Bangladesh Garment Sector". DECRG, World Bank, mimeo. Lezama, M., B. Webber, and C. Dagher. 2004. Sourcing Practices in the Apparel Industry: Implications for garment exporters in Commonwealth Developing Countries. The Commonwealth Secretariat, London. Manole, V. 2005. "Winner or Loser? Effects o f Quota Abolition in World Markets for Textile and Apparel". DECRG, World Bank, mimeo. . 2005. "Trade Liberalization for Textiles and Apparel: The Case of Bangladesh". DECRG, World Bank, mimeo. Mlachila, M., and Y. Yang. 2004. The End of Textiles Quotas: A Case Study of the Impact on Bangladesh. Worlung Paper. IMF, Washington. MinistryofCommerce. 2001. Readymade Garment Sector in Bangladesh: Background Information for a Strategy Study. Government o f Bangladesh, July. . 2004.TradeReform and Export Response: Post-MFA Policy Implicationfor Bangladesh. Government of Bangladesh, June. Martin, W., Manole, V., and Van Der Mensbrugghe, D., 2004, "Dealing with Diversity: Analyzing the Consequencesof textile quota abolition", presented at GTAP Conference, June 2004, Washington D.C. Moulton, Brent R. 1990. "An Illustration of a Pitfall inEstimating the Effects of Aggregate Variables on Micro Unit." Review of Economics and Statistics. 334-338. National Co-ordination Council on RMG Sector of Bangladesh. 2004. Post-MFA Challenges of RMG Sector of Bangladesh: Review, Recommendations and Action Plan. Dhaka, July. Nordas, H. K. 2004. The Global Textile and Clothing Industry Post the Agreement on Textile and Clothing.Discussion Paper. WTO, Geneva. OECD. 2004. A New WorldMap in Textileand Clothing: Adjusting to Change. Paris. 56 6 Olley, G. Steven and Anel Pakes. 1996. "The Dynamics of Productivity in the Telecommunications EquipmentIndustry." Econometrica Vol. 64 (6): 1263-1297. Paul-Majumdar P. (2005) "Bangladesh RMG Moving Towards Informal sector". BIDS Annual Journal, V O l . 22. Pursell, G. 2005. Free Trade Between India and Bangladesh? A Case Study of the Readymade Garment Industry.Mimeo. World Bank. Razzaque, A. 2005. Sustaining RMG Growth after MFA-Phase-out: An Analysis of Relevant Issues with Reference to Tradeand Human Development. UNDP, Dhaka, January. Suhrawardy G.M., Ghosh, P.S. and Hossain, M. A. (2004), "Post MFA - Potential Disempowerment of RMGwomen workers". Bangladesh Economic Association Conference, December 8-10,2004 Taslim, M.A. 2004. Regional Cumulation: Implications for Textile and RMGIndustries. Mimeo. Bureau of Economic Research, September. United States International Trade Commission. 2004. Textile and Apparel: Assessment of the Competitiveness of CertainForeign Suppliers to the USMarket. USITC Publication3671, January. World Bank. 2003. Textileand ClothingPolicy Note: Implicationsfor Pakistan of Abolishing Textileand ClothingExport Quotas. Policy ResearchWorkmg Paper. World Bank, Washington, D.C. World Bank, 2004. TradePolicies in SouthAsia: An Overview. ReportNo. 29949. Washington D.C. . 2005a.Bangladesh: Growth and Export Competitiveness. South Asia Poverty Reduction and Economic Management. Washington D.C., April. . 2005b. Winner or Loser? Efects of Quota Abolition in World Markets for Textile and Apparel, mimeo. DECRG Trade, Manole V, 2005. WTO (2005) `Optionsfor Least-Developed Countries to Improve their Competitiveness in the Textiles and ClothingBusiness' WTlCOMTDILDClW37 57 Annexes 59 Annex A GTAPComputableGeneralEquilibriumModeland itsLimitations To determine the global and country-by-country effects of the abolition o f MFA quotas, a computable general equilibrium (CGE) model formulated under the GTAP (Global Trade Analysis Projects) was used after adequately restructuring the model to cover and as correctly as possible portray the following: 0 Firstthe highlycomplex interactionbetween tariff reductions and quota liberalization on the one side and changes intrade flows and key economic indicators on the other. 0 Secondly, Greater China's accession to the WTO, with all its implications for the global textile and clothing industries and trade in general i s explicitly incorporated into the exercise. 0 Thirdly, exploring the implications o f China's policy o fexport restraint through the imposition o f export tax. 0 Fourthly, determining the implication for Bangladesh o f duty-free access for apparels in the U S market. The main thrust o f the exercise was o f course to determine the likely impact o f carrying out the ATC and eliminating quotas on T&C exports. Insimple terms this means that we examine the changes which will occur with respect to output, foreign trade and income when the ATC has been effected and when Greater China's WTO accession has been completed with respect to China (PRC) as well as Chinese Taipei. The "computable general equilibrium" (or CGE) model used here allows assessment o f the economic impact o f regional, multilateral and global trade agreements. It likewise permits the assessment o f liberalization across broad sectors o f individual economies, including interactions between sectors that may result. The estimated effects from the CGE model at the national level, o f course, reflect the interactions with neighboring economies as well as with economies/regions in other parts o f the world. SomeBackgroundofthe Model The GTAP 6.4 version o f the model used belongs to a family o f economic models characterized by an input-output structure (based on regional and national input-output tables) that explicitly links industries in a value added chain from primary goods, over continuously higher stages o f intermediate processing, to the final assembling o f goods and services for consumption. Linkages between sectors are both direct (like the input of textiles inthe production o f automobiles) and indirect (like the use o f mining inputs into steel, which feeds into machines which weaves the textiles). The model captures these linkages by modeling firms' use of factors and intermediate inputs when producing goods and services. The most important aspects o f the model can be summarized as follows: 0 it covers allworld trade andproduction; 0 it includes intermediate linkages between sectors; 0 and it allows for trade to affect capital stocks through investment effects, hence.we model medium to long-run investment effects. The DataUsedinthe Model The data come from a number o f sources. Data on production and trade are based on national accounting data linked through trade flows and drawn directly from the Global Trade Analysis Project (GTAP)version 6.4 dataset. The GTAPversion 6.4 dataset is benchmarked to 2001, and includes detailed 61 national input-output, trade, and final demand structures. Significant modifications have been made to the basic GTAP database. The basic social accounting and trade data are supplemented with trade policy data, including additional data on tariffs and non-tariff barriers. The dataset has also been updated to better reflect actual import protection for goods and services (the basic GTAP database includes no informationat all on trade bamers for services). SomeLimitations ofthe Model Since this exercise is based on an economic model, it i s useful - as with all models - to keep the limitations o f the exercise in mind. The GTAP model i s a static, multi-country and multi-sector global general equilibrium model capable o f generating comparative-static simulations; (one could mimic some dynamic linkages, but fully dynamic solutions are not possible). By necessity, the production structure i s based on the inclusion o f highly `aggregated' sectors. Thus, for instance, clothing and textiles are each treated in the model as single, homogenous products or fairly aggregated activities (despite the fact that there i s a wide range o f apparel items, not all of which were subject to quotas). This (very understandable) limitation on the level o f disaggregation in modeling the production side would o f course affect the simulationresults on the likely (comparative static) impacts by ignoring intra-sectoral specializations. --On the demand side, the GTAP 6.4 version has attempted to model product heterogeneity by treating T&C exports o f competing countries as differentiated products, which would be certainly an important effort to improve the reliability o f simulations results. --On the supply side, because o f the model's highly aggregated production structure, the simulations results would not fully capture some o f the positive resource allocation shifts that would be induced in each country, including Bangladesh, towards relatively more competitive and profitable individual items--i.e., the induced intra-sectoral production and trade specialization (among hundreds and hundreds o f differentiated apparels and textiles products) in the post-ATC period. This would meanthat the current results mightbe underestimating positive output/export impacts (or over-estimating adverse impacts). Not considered are the likely dynamic,long-term positive effects o fthe removal o f the T & C export quotas, though this might be another point that could be made on the side o f the likely long-termpositive impacts. Finally, one can think o f some generalized limitations associated with most models that are designed to approximate rather than replicate reality. First, the model cannot forecast all future events. It i s highly likely that unanticipated economic, political, andor natural events will occur and will have important effects on some o f the agents and activities identified inthe model. (Consider, for example, the impact o f the East Asian financial crisis, which was not envisaged in the ATC formulation). In this regard, one could think o f the model as saying "in a world like the one we currently observe and with the assumed demand and production structure, ifpolicies were different, this world would then change inthe ways reportedinthe tables." Thus the model makes no claim o fbeing able to forecast future changes with precision. Rather, in the absence o f surprises (which no one can predict with any degree o f certainty for the next decade), the model helps to generate estimates o f likely economic effects. Another limitation is the simplifications embodied in the model. This poses a standard challenge. When economic policy and its impacts are being modeled, the strategy i s to develop a reasonable, though stylized representation o f complex policy, demand, and production relationships. The trade-off is between keeping the model workable, and keeping it realistic enough to actually be useful. This having been said, it is worth noting that this class o fmodels does actually do well inidentifying resource, production, and trade shift 62 Annex B Economics of Quota The imposition o f a quota restricts access of exports into a market. If the quota i s binding, the price o f the restricted product will rise in this market, but it could fall in any non-restricted markets. Quotas, therefore, have price (and therefore, quantity) implications not only inthe quota restrictedmarket but also in non-restricted markets. This can be illustrated by Figure 1 due to Kathuria, Martin and Bhardwaj (2001). The demand curves, Di, are drawn as downward-sloping by adopting the Armington assumption. The subscript Urefers to the unrestricted market, R to the restricted market and T to the total demand. Inthe absence o f any restrictions, the global price will be pw.At this price pwAand pwBwill be sold in the two markets respectively. Ifa quota equal to OQ i s imposed, the effective global demand curve i s given by D'T(Figure 2). The country now exports an amount equal to the quota to the restricted country (assuming the quota i s binding) at a price pR,and exports puE to the unrestricted country at price pu. The imposition of the quota raises the price of exports inthe restricted market, but reduces it inthe unrestricted market. The quota reallocates some o f the exports to the unrestricted market and hence the price has to be reduced to absorb the additional amount. The abolition o f the quota will reverse the process. The price o f the product in the unrestricted market will rise with a concomitant fall in quantity. On the other hand, the price inthe quota-restricted market will fall with an accompanying increase inthe quantity sold. Exporters will shift part of their sales to the highprofit quota-restricted market by reducing the sales inthe unrestricted market. Figure 2 provides an estimate o f the quota premium. For there to be an equilibrium, the quota premium must be such as to exhaust any advantage o f selling inthe restricted market. Hence, the quota premium will settle at the difference between the prices in the two markets (pR-pu). In other words, the quota-holder would be indifferent between producing and selling the product in the restricted market at price pR,or cede his right to others for a quota premium o f (pR-pu). Alternatively, a producer who does not hold quota will be willing to pay up to (pR-pu) in order to acquire the quota and sell inthe restricted market. Ifthere i s a free market for quota transactions, the quota premium will tend toward (pR-pu). Of course, the actual quota price may be less than this amount. There may be transaction costs that reduce the margin. Large importers and retailers may use their market power to extract a slice o f the quota premium such that the quota-holder may not receive the full market premium. Abstracting from these qualifications, the quota premium, provides a good measure o f the excess o f market price over cost. This i s also a measure o f the competitive strength of the country as the market price could be reduced by t h s margin and still maintain production at the quota-restricted level. The effect o f the quota can be duplicated inan otherwise free market if a tax equal to the quota premium i s imposed on the export of the pr~duct.~'Hence, the analysis i s also referredto as export tax equivalent approach. The analysis above explains the reallocationof exports o f the RMG sector o f the country between its quota-restricted and unrestricted markets in a static framework. It does not indicate how the market shares o f the competing countries will be affected by the abolition o f the quota regime. T o understand the market share implications o f quota-abolition we use Figure 3. Let the line marked C be the export supply curve o f a particular apparel item o f a (group of) low-cost country C,the line market B the export supply curve o f a (group of) moderate cost country, B, and the line market Ebe the export supply o f the item o f a (group of) high cost country, E. The high cost country in this example may also be regarded as the domestic economy. The line E then represents the supply curve o f the domestic firms o f the economy. The line marked D i s the domestic economy's demand for the item. Horizontal summation o f the supply curves o fthe three countries gives the aggregate supply curve amnSF under free market conditions. 57 See Kathuria and Bharadwaj (1998). 63 N o w assume that both the low cost and the moderate cost countries are quota-restricted. Each o f them can sell at most Qc amount o f the product. In order to avoid cluttering up the figure too much, the supply curves and quota amounts are so drawn that the quotas are binding. As a result o f the quotas, the supply function is no longer am&, but the step function abvdeSR. As long as the price is less than b, only the low cost country supplies the market. At price b, the low-cost country hits the quota ceiling Qc. As the prices increase to v, the profitability o f exporting Qc increases but export o f the low cost country cannot expand. When the price exceeds v, the moderate cost country can enter the market. Its export increases along vd until the price reaches d at which point it, too, encounters a quota ceiling. As the price increases further, the quota-rent increases for both countries, but export supply remains constant at Q. When the price exceeds e, the high cost producer E can enter the market. As E is not quota restricted any increase inprice beyond e raises the output supplied to the domestic market along eSR. Hence, the quota distorted supply curve for the item i s abvdeSR. The intersection o f this supply curve and the demand curve D establishes PRas the equilibrium price and R as the equilibrium quantity in the quota-restricted market. Each o f the restricted countries supplies Qc amount o f the itemwhile the rest (R-Q=Pu) i s supplied by the highcost country. When the quota regime i s abolished, the relevant aggregate supply curve i s armsF. Its intersection with the demand curve D establishes PFas the free market price and F as the total purchase o f the economy. Since the free market supply curve lies below the quota-distorted supply curve, price PFR. The abolition o f quota unambiguously raises the market size, but reduces the price. Whether the total purchase F in value terms will increase or not will o f course depend on the elasticity o f demand. Ifdemand i s elastic (as i s most likely for apparels), total export invalue terms will also increase with an increase inexport volume (PFF >PRR). At the new unrestrictedprice, the export o fthe most competitive country increasesmost, from QC to C'. The moderate cost country B also increases its export, though not by as much as the low cost country. Its export increases from Qc to B'. The high cost country suffers from the change; its production (export) declines from (R-Q)to only E'. The reallocation o f production and export after the abolition o f the quota regime unambiguously increases global welfare (in terms o f an increase in consumer and produces surplus) but reduces apparel production in the high cost country imposing some adjustment costs on it. The severity o f the cost will obviously depend on the magnitude o f the adjustment. The diagrammatic analysis above provides an indication o f what to expect in the post-MFA era. It also suggests the conditions under which the predictionswill be different from what have beendepicted above. It i s certain that the abolition o f quotas will reduce the market price so long as some o f the supplying countries were quota-restricted. It is also certain that the low cost country will gain from quota abolition - its output and export will expand. The high cost country on the other hand will witness a reduction i s production and/or export. However, what happens to the country in-between will depend very much on the cost and demand conditions. For the configuration o f the supply and demand curves as drawn above the moderate cost country also gains from quota abolition. But it i s easy to see that the lower the cost o f production inthe low cost country, i.e. the lower the supply curve C, the smaller would be the gain to the moderate cost country; and beyond some point it will actually suffer a reduction in output and export. The reverse also holds. The problem (and prospect) o f the RMG sector o f Bangladesh in the post-MFA world will be evident ifinthe diagram above the low cost country i s interpreted to be a country such as China or India, the high cost country to be USA, and the moderate cost country to be Bangladesh. Whether Bangladesh will fare well inthe emerging market situationwill depend on the competitiveness o fthe industry, i.e. the position and the slope o f its export supply curve. If it can bring down its cost at a faster rate than the lower cost countries, i.e. if it i s dynamically efficient, it can actually increase the market share. Note that 64 even ifBangladesh were to gain initially from quota abolition, as indicatedby the figure above, it will not be able to hold on to the gains unless it can maintain the competitive position. If other countries are becoming more efficient through industry restructuring and rationalisation, appropriate government polices, superior infrastructure and importantly technological innovation, Bangladesh will have to respond with appropriate policies and actions in all these areas by both the government and the private sector inorder to remain competitive and maintain market share. 0 Quantity F i p2, Madsetequilibriuminthe presenceofquotas PR p* Pu 65 Figure3. Market Equilibriumafter quotaabolition 66 Annex Table 1: Unit Price of Selected Apparel Items to USA 2002 2003 Yearto 2002 2003 Year to date date 10104 10104 237: Playsuit/sunsuit($ per dozen) 331: Cottongloves/mittens($ per DPR) World 32.74 33.16 28.54 World 2.30 2.17 2.00 Bangladesh 30.83 36.08 39.93 Bangladesh 0.84 0.68 0.76 China 28.31 29.81 25.67 China 2.04 1.84 1.76 India 42.68 48.75 42.26 India 1.68 0.64 0.87 Indonesia 42.55 41.53 43.90 Indonesia 1.09 0.95 1.14 Mexico 56.31 50.60 42.04 Mexico 3.97 4.01 3.03 Pakistan 20.96 23.48 25.34 Pakistan 1.74 1.60 1.70 Sri Lanka 54.42 49.10 46.66 Sri Lanka 3.76 3.15 3.28 Vietnam 30.16 25.95 33.10 Vietnam 2.09 2.39 2.19 334: Other coats, M/B ($ per dozen) 335: Cotton coats, W/G ($ per dozen) World 118.66 117.18 121.50 World 108.84 107.97 117.08 Bangladesh 92.82 90.39 101.89 Bangladesh 76.60 86.96 86.58 China 165.22 185.03 179.89 China 192.26 208.76 228.05 India 103.58 113.70 122.07 India 94.54 110.42 110.59 Indonesia 129.29 125.59 116.67 Indonesia 128.05 131.99 132.10 Mexico 185.20 224.46 215.07 Mexico 89.58 97.50 105.13 Pakistan 70.58 70.72 81.05 Pakistan 49.26 52.76 50.84 Sri Lanka 95.10 110.23 132.25 Sri Lanka 98.68 103.55 100.65 Vietnam 76.84 99.68 144.75 Vietnam 62.84 75.78 101.45 336: Cottondresses ($ per dozen) 636: Dresses, MMF($ per dozen) World 61.32 60.10 60.67 World 107.74 108.19 106.60 Bangladesh 42.93 45.70 39.58 Bangladesh 48.53 48.99 45.00 China 113.38 132.29 167.95 China 309.61 339.59 397.22 India 70.66 70.42 57.57 India 79.29 83.11 77.62 Indonesia 62.51 58.82 67.37 Indonesia 84.24 80.73 72.22 Mexico 53.24 55.32 62.85 Mexico 65.52 61.58 64.12 Pakistan 24.27 21.51 21.54 Pakistan 32.75 30.96 32.45 Sri Lanka 65.10 65.37 66.99 Sri Lanka 88.03 88.64 89.77 Vietnam 47.59 46.22 42.36 Vietnam 70.64 68.70 66.10 338: Knit shirts,M/B($ per dozen) 339: Knit shirts/blouses,W/G ($ per dozen) World 35.95 33.24 33.49 World 38.82 37.00 36.39 Bangladesh 44.87 42.57 39.92 Bangladesh 44.25 43.54 37.77 China 83.64 87.96 78.52 China 61.80 64.54 74.03 India 61.45 67.40 65.10 India 57.22 62.17 58.18 Indonesia 75.42 70.29 70.10 Indonesia 69.17 62.20 55.66 Mexico 26.99 25.52 25.00 Mexico 28.39 26.43 23.98 Pakistan 48.39 45.01 45.27 Pakistan 36.50 34.75 30.16 Sri Lanka 81.79 80.92 73.56 Sri Lanka 52.60 52.36 51.99 Vietnam 44.48 43.25 48.54 Vietnam 35.09 36.00 37.16 340: Non-knit cottonshirts,M/B($ per dozen) 640: Non-knit MMFshirts, M/B($ per dozen) World 74.22 74.71 78.18 World 52.11 52.30 51.91 Bangladesh 53.06 50.07 50.89 Bangladesh 55.37 50.50 45.04 China 76.60 79.99 85.92 China 49.62 51.13 51.62 India 59.36 64.11 71.92 India 47.79 56.72 68.26 67 Indonesia 70.08 73.43 75.78 Indonesia 63.36 64.20 62.70 Mexico 73.52 80.57 79.71 Mexico 66.67 63.94 62.27 Pakistan 40.20 35.68 38.42 Pakistan 31.21 26.17 19.50 Sri Lanka 71.36 66.06 64.43 SriLanka 30.36 60.11 51.46 Vietnam 48.12 50.01 58.94 Vietnam 46.97 46.77 56.26 341: Non-knit cotton blouses,W/G (%per dozen) 342: Cotton skirts (%per dozen) World 61.60 63.18 64.58 World 67.68 65.81 64.91 Bangladesh 39.05 36.94 39.94 Bangladesh 50.10 53.98 54.76 China 100.13 117.78 109.02 China 103.59 125.24 135.52 India 52.78 57.98 52.09 India 58.65 65.90 62.46 Indonesia 64.04 67.10 74.63 Indonesia 72.67 82.90 81.63 Mexico 51.95 60.87 59.91 Mexico 80.22 72.60 56.15 Pakistan 33.00 33.43 36.88 Pakistan 36.83 35.06 34.57 Sri Lanka 57.25 58.15 58.84 Sri Lanka 58.49 60.84 63.74 Vietnam 39.79 44.10 54.32 Vietnam 50.85 47.62 53.25 642: Skirts, MMF(%per dozen) 347: Cotton trousers, M/B(%per dozen) World 71.41 69.74 69.35 World 77.65 77.29 78.38 Bangladesh 55.53 62.33 58.39 Bangladesh 60.68 63.89 70.08 China 100.37 113.06 118.58 China 105.70 104.80 118.61 India 64.49 67.54 71.78 India 99.02 101.93 115.59 Indonesia 70.18 75.96 73.44 Indonesia 79.48 88.49 87.34 Mexico 46.29 48.75 45.62 Mexico 87.70 89.00 88.98 Palustan 28.87 28.07 27.46 Pakistan 63.20 62.07 67.99 Sri Lanka 62.06 62.46 59.00 Sri Lanka 79.64 94.92 88.61 Vietnam 44.70 53.55 51.39 Vietnam 57.38 59.82 72.20 348: Cotton trousers, slacks, etc. W/G (%per dozen) 351: Cotton nightweadpyjamas (%per dozen) World 71.45 70.57 73.52 World 52.47 48.15 48.70 Bangladesh 60.05 64.51 67.56 Bangladesh 41.51 44.27 38.11 China 117.69 118.70 135.80 China 78.99 76.96 79.44 India 100.04 107.37 102.01 India 73.58 72.39 63.92 Indonesia 79.96 95.06 87.25 Indonesia 51.38 52.79 55.67 Mexico 78.38 84.17 86.04 Mexico 54.94 48.49 42.45 Pakistan 50.23 55.44 62.97 Pakistan 45.98 39.25 37.37 Sri Lanka 72.40 81.41 77.51 Sri Lanka 70.04 72.68 68.80 Vietnam 52.91 51.35 62.95 Vietnam 32.82 34.38 66.56 651: Nightweadpyjamas, MMF($ per dozen) 352: Cotton underwear (%per dozen) World 47.97 44.69 44.74 World 12.04 11.57 11.25 Bangladesh 43.81 44.36 45.25 Bangladesh 7.68 7.95 8.11 China 77.17 70.20 74.68 China 30.58 29.52 33.49 India 75.68 37.24 61.35 India 12.50 13.24 13.47 Indonesia 57.58 67.08 58.51 Indonesia 23.31 22.67 19.57 Mexico 46.34 46.45 46.62 Mexico 15.09 15.51 14.88 Pakistan 41.27 35.95 37.05 Pakistan 14.84 15.97 16.23 Sri Lanka 75.22 79.10 73.89 Sri Lanka 22.82 27.04 23.23 Vietnam 33.87 41.89 45.90 Vietnam 8.56 9.74 11.67 652: Underwear, MMF(%per dozen) 363: Cotton terry and other pile towels (%per unit) World 15.62 15.79 15.87 World 1.61 1.70 1.81 Bangladesh 7.31 6.69 7.66 Bangladesh 0.43 0.57 0.74 China 19.40 18.83 21.90 China 2.68 2.98 3.34 India 12.70 8.40 19.40 India 1.55 1.82 1.92 Indonesia 14.79 20.03 11.26 Indonesia 1.51 1.69 2.63 68 Mexico 20.56 19.25 16.42 Mexico 1.05 1.89 2.48 Pakistan 25.73 20.29 9.93 Pakistan 1.13 1.27 1.59 Sri Lanka 33.33 38.74 34.71 Sri Lanka 0.56 0.64 0.77 Vietnam 2.72 5.23 5.89 Vietnam 0.61 0.89 0.97 369: Other cottonmanufactures($ per kg) 634: Other coats, MMFM/B($ per dozen) World 5.86 5.63 5.70 World 121.57 123.00 124.57 Bangladesh 2.57 2.44 2.58 Bangladesh 99.95 103.87 106.03 China 7.59 6.68 6.59 Chma 227.03 259.29 282.59 India 5.18 5.14 5.29 India 115.33 115.21 121.89 Indonesia 7.35 5.70 4.57 Indonesia 219.58 226.85 169.00 Mexico 7.47 8.14 6.12 Mexico 87.76 93.88 108.11 Pakistan 3.17 3.14 3.37 Pakistan 60.23 63.24 60.24 Sri Lanka 3.09 2.32 3.04 Sri Lanka 110.56 114.03 114.95 Vietnam 3.97 6.78 8.29 Vietnam 106.25 123.15 123.50 635: Coats, MMFW/G ($ per dozen) 638: Knit shirts, MMFMIS($ per dozen) World 116.02 117.03 116.50 World 34.10 35.57 36.50 Bangladesh 93.44 103.61 115.08 Bangladesh 32.71 30.84 27.15 China 231.24 251.23 258.44 China 66.34 72.33 75.83 India 105.44 94.83 97.20 India 41.25 40.43 41.82 Indonesia 167.18 172.76 131.21 Indonesia 42.29 42.07 40.63 Mexico 76.81 80.60 79.34 Mexico 29.15 27.91 28.71 Pakistan 43.58 46.20 40.46 Palustan 22.59 22.88 22.56 Sri Lanka 107.67 110.76 112.06 Sri Lanka 52.14 50.43 5 1.05 Vietnam 118.22 113.78 108.09 Vietnam 38.22 43.08 55.90 639: Knit blouses, MMFW/G ($ per dozen) 641: Non-knit shirts & blouses, MMFW/G ($ per dz) World 53.95 53.56 53.42 World 56.71 55.11 51.78 Bangladesh 31.98 33.83 36.20 Bangladesh 42.31 43.05 40.12 China 80.98 78.20 80.18 China 80.58 76.66 75.34 India 38.25 39.33 42.66 India 44.70 44.75 47.73 Indonesia 51.70 52.74 55.25 Indonesia 48.72 48.09 45.46 Mexico 41.40 43.39 42.34 Mexico 44.90 46.24 42.84 Pakistan 29.03 23.34 22.60 Pakistan 25.65 25.05 24.35 Sri Lanka 50.66 50.93 46.94 Sri Lanka 61.36 62.88 , 56.02 Vietnam 42.81 44.71 49.44 Vietnam 39.38 45.78 49.67 645: Sweater,MMFMIS ($ per dozen) 646: Sweater, MMFW/G ($ per dozen) World 61.73 65.99 58.74 World 68.36 61.62 61.69 Bangladesh 42.35 49.89 37.16 Bangladesh 40.18 32.29 31.59 China 68.25 77.83 65.95 China 98.09 79.71 85.87 India 56.50 51.37 69.02 India 81.56 79.35 91.50 Indonesia 50.44 53.27 45.01 Indonesia 60.36 53.14 48.73 Mexico 70.62 85.26 72.54 Mexico 65.03 66.80 92.01 Pakistan 23.61 Pakistan 45.10 21.20 Sri Lanka 64.92 66.11 68.15 Sri Lanka 68.92 71.98 68.82 Vietnam 37.64 42.31 52.03 Vietnam 64.20 58.82 61.08 647: Trousers,breeches, shorts, MMFMIS($ per doz) 648: Trousers, breeches, shorts, MMFW/G ($ per doz) World 55.64 55.47 55.19 World 59.01 59.81 58.69 Bangladesh 40.12 39.53 36.50 Bangladesh 46.94 52.07 5 1.59 China 76.23 84.98 97.76 China 96.28 104.13 106.44 India 51.48 51.40 52.93 India 55.01 45.40 43.23 Indonesia 53.30 53.41 51.28 Indonesia 62.09 64.83 62.76 Mexico 56.38 55.71 56.54 Mexico 46.87 49.54 47.29 Pakistan 31.52 27.94 27.72 Pakistan 33.17 22.46 19.88 69 Sri Lanka 58.42 58.65 49.35 SriLanka 67.20 64.85 61.13 Vietnam 45.85 48.87 69.81 Vietnam 43.05 45.16 61.77 847: Trousers, etc. Silk & Vegetable ($ per dozen) World 75.07 64.00 60.51 Bangladesh 40.33 36.98 36.70 China 73.81 62.93 59.59 India 95.50 94.36 81.71 Indonesia 49.26 40.65 45.76 Mexico 103.88 84.40 124.70 Pakistan 35.73 35.41 28.36 Sri Lanka 114.20 110.44 93.37 Vietnam 46.64 71.34 64.23 Source: Office ofTextiles and Apparel, U S Department of Commerce (www.otexa.ita.doc.gov) 70 Annex Table 2: Average quota prices of selected items in selected countries QUOTA: Averageprice (US$) Cat Unit Pakistan India Indonesia China Bangladesh Description 2003 2001 2002 2002 237 DOZ Play Suits, Sun suits 0.01 0.02 2.50 1.06 3341634 DOZ Other coat MB 5.99 7.15 7.90 37 13.77 3351635 DOZ Coat W&G 0.20 1.79 7.90 41 11.61 3361636 DOZ Dresses 0.56 16.09 2.26 33.5 3.76 338 DOZ Knitted Shirts M & B 8.70 16.09 5.43 46.5 11.56 339 DOZ KnitShirts & 6.10 16.09 5.75 46.5 Blouses W & G 3401640 Doz 20.94 1.63 26 3.41 347148 DOZ Trousers, Slacks & 41.60 46.72 10.18 39 21.65 Shorts M&B 3511651 SME Nightwears and 5.85 15.32 5.5 20.75 5.08 Pajamas 3521652 DOZ Under wears 3.68 5.5 0.40 6341635 other coats MMF 39.63 40.5 M/BandWIG 6381639 DOZ Knit Shirts, M& B 3.71 1.46 22 1.62 640 Doz 1.77 641 Doz W&GShrts& 2.30 15 0.36 Blouses, not Knit 6451646 Doz Sweaters 23.5 6471648 DOZ M/B, WIG Trousers, 0.48 12.77 2.66 16.25 9.3 1 Breeches & Shorts Mean 6.99 15.52 7.08 27.7 6.97 Source: Communicationtexwatch.com 71 Annex Table3: Unitpricesof garmentsexported from Bangladeshand China, 2003 Cat. Product China Bangladesh China China- Bangladesh Bangladesh quota quota FOB net FOB net price ratio price price o f quota of quota price price US$ US$ US$ US$ 237 Playsuits, etc. 0.00 2.86 25.60 50.22 2.0 331 Gloves 0.00 1.14 7.07 15.68 2.2 334 Coats, non-suit, 0.00 42.29 85.98 157.02 1.8 M&B 335 Coats, W&G 0.92 43.43 61.28 155.35 2.5 3361636 Dresses 0.86 38.29 47.52 354.50 7.5 3381339 Knit Shirt& 1.22 53.15 43.52 33.68 0.8 Blouses 3401640 Shirts, not Knit, 0.27 29.72 44.76 40.03 0.9 M&B 341 Shirts & Blouses, 0.08 30.86 28.15 82.79 2.9 not Knit, W&G 3421642 skirts 2.65 38.29 53.01 91.02 1.7 3471348 Trousers, etc. 6.63 44.58 66.27 99.19 1.5 3511651 Nightwear 0.99 23.72 39.64 57.38 1.4 3521652 Underwear 0.02 6.29 7.48 17.53 2.3 363 Terry & otherpile 0.00 0.00 0.50 3.12 6.2 Towels 369-S* Shop Towels 0.00 0.00 10.24 0.00 634 Coats, non-suit, 2.48 42.29 99.26 265.99 2.7 M&B 635 Coats, W&G 2.60 50.29 103.84 208.15 2.0 6381639 Knitshlrts & 0.15 25.15 25.44 58.39 2.3 Blouses 641 Shirts & Blouses, 0.18 17.15 30.24 68.79 2.3 not Knit, W&G 6451646 Sweaters 0.51 26.86 28.38 75.91 2.7 6471648 Trousers, etc. 18.57 89.31 Source: Eurostat (http:lleuropa.eu.int/commleurostat) 72 Statistical Annex 73 - Ta 1: Bai ladesh acroec - - - omic 1 icator! - 'y05 lescription -Ave FY04 1980s FY92 FY96 FY97 FY98 FY99 FYOO FYOl FYO2 --0 'YO3 Srowth Rates (YO) 3DP Growth 3.7 5.0 4.6 5.4 5.2 4.9 5.9 5.3 4.4 5.3 6.3 5.4 3DP Growth Per Capita 1.2 3.0 2.8 3.6 3.5 3.6 4.9 3.7 2.9 4.0 6.3 5.4 'er Capita GDP Atlas Method 347.9 350.7 354.1 367.3 371.1 372.0 389.0 418.0 145.0 US$) 226.6 298.0 332.8 Saving & Investment :Yoof GDP) 3 0 s s Domestic Saving 11.6 13.9 14.7 15.9 17.4 17.7 17.9 18.0 18.2 18.6 19.5 20.2 3oss National Saving 17.0 19.3 20.0 21.6 21.8 22.3 23.1 22.4 23.4 24.9 25.4 26.5 ?rivate Investment 11.3 10.3 13.6 13.7 15.3 15.5 15.6 15.8 16.8 17.2 17.8 18.5 ?ublic Investment 5.5 7.0 6.4 7.0 6.4 6.7 7.4 7.2 6.4 6.2 6.2 5.9 Central God.Budget [Yoof GDP) rota1Revenue 8.8 8.3 9.0 9.2 9.3 9.0 8.5 9.0 10.1 10.3 10.1 10.6 Total Expenditure 17.2 12.7 13.4 13.5 13.3 13.8 14.7 14.1 14.8 13.7 13.3 15.1 3vmall BudgetBalance 8.3 4.5 4.5 4.3 4.1 4.8 6.2 5.0 4.6 3.4 3.2 4.5 Balance of Payments (Yo of GDP) Exports 4.3 6.4 9.5 10.5 11.7 11.6 12.2 13.8 12.5 12.6 13.3 12.8 Imports 12.4 -11.3 -17.1 -16.9 -17.1 -17.5 -17.8 -19.9 -16.3 -16.8 -17.4 -19.0 Services & Income (net) -0.4 -0.1 -0.1 0.1 0.2 0.1 -0.1 -0.5 -1.7 -1.7 2.2 2.5 Current Transfers 4.1 4.6 4.5 5.1 4.6 4.9 5.7 4.9 6.0 6.7 6.6 6.9 Current Account Balance -0.6 -0.9 0.0 -1.7 0.5 0.6 0.3 -1.8 ( including transfers) -3.9 -0.4 -3.2 -1.3 External Indicators External Debt (US$b.) 7.9 13.3 15.2 15.0 14.0 14.8 16.2 15.1 16.3 16.5 16.8 17.8 Ext. Debt as % o f GDP 34.2 39.5 37.3 34.7 31.6 32.7 34.0 30.8 34.4 31.9 29.5 29.3 BBGross Reserves 1.8 1.5 1.6 1.3 1.6 2.5 2.7 2.7 (US$b.) (end of period) 0.5 1.6 2.0 1.7 BB Gross Reserves (in 2.8 2.3 2.3 1.7 2.1 2.9 2.8 2.4 months o f imports) 0.7 5.5 3.5 2.9 External Debt Service Ratio 7.9 8.4 8.0 6.4 6.1 5.6 4.3 5.0 (% of Export Earning) 7.8 15.8 10.7 9.6 Exchange Rate Nominal Period Average (TK/US$) 26.8 38.2 40.8 42.7 45.5 48.1 50.3 54.0 57.4 57.9 58.9 60.5 Nominal End o f Period (WUS$) 34.9 39.0 41.8 43.7 46.3 48.5 51.0 57.0 57.9 58.5 60.4 63.7* Real Effective (1990=100) 107.0 95.4 93.4 93.5 101.9 104.4 102.6 100.3 97.5 93.6 Rate of Inflation 1.9 2.8 4.4 5.8 6.5 (YO)(year on year) 10.8 4.6 6.8 2.5 8.7 7.1 2.8 Total Public Debt (YOof GDP) N A NA 45.7 43.8 41.5 43.8 46.9 45.2 52.7 51.0 48.3 47.9 MemorandumItems GDP at Current. Prices 2,535.2 2,732.0 3,005.8 3,329.; 3,684.8 (Taka bill.) 619.8 I,I 1,663.2 95.4 1,807.0 2,001.8 2,197.0 2,370.9 GDP at Current. Prices Atlas 22.5 33.8 40.6 43.2 44.4 45.4 47.7 48.9 49.5 51.7 56.5 61.0 Method (US$bill) Population (mill.) 98.7 113.0 120.8 122.6 124.5 126.3 128.1 129.9 131.6 133.4 135.2 137.0 Population growth Rate 2.5 1.4 1.7 I.5 1.5 1.4 1.4 1.4 1.3 1.4 1.3 1.3 Source: Various publications of the World Bank, ADB and Bangladesh Bureau o f Statistics *As o f 19 June 2005 The Atlas Method was not used for calculating average per capita GDP and GDP at current market prices in the 1980s, in FY04 and FY05 75 Table 2: Bangladesh Balance of Payments (i nillion L. 9 I Items FY98 FY99 FY2000 FYOl FY02 FY03 FY04 BY05 :estimate) projection) -1,669 -1,934 -1,865 -2,011 -1,768 -2,215 -2,319 -3,783 5,103 5,283 5,701 6,419 5,929 6,492 7,521 7,827 -6,772 -7,217 -7,566 -8,430 -7,697 -8,707 -9,840 -11,610 -570 -603 -645 -914 -499 -691 -814 -1,141 -100 -135 -221 -264 -319 -458 -374 -354 1,876 2,195 2,394 2,171 2,826 3,440 3,743 4,200 126 220 165 72 69 82 61 52 1,750 1,975 2,229 2,099 2,757 3,358 3,682 4,148 1,525 1,706 1,949 1,882 2,501 3,062 3,372 3,824 -463 -477 -337 -1,018 240 76 176 -1,078 445 387 561 432 410 428 319 309 445 387 561 432 410 428 319 309 FinancialAccount 237 -395 -185 249 71 581 173 845 Direct Investment 249 198 194 174 65 376 385 410 Portfolio Investment 3 -6 0 0 -6 2 6 0 Other Investment -15 -587 -379 374 12 634 242 931 MLTloans'/ 706 821 806 790 733 1,070 734 1,373 -308 -341 -396 -416 -421 -436 -492 -442 ' MLT amortizationpayments Other long term loans (net) -47 -41 127 -13 -42 -20 -41 -45 ' Other short term loans (net) 168 -78 56 86 20 142 13 150 Other assets -41 -58 -55 -68 -52 -125 -125 -200 Trade credit (net) -522 -829 -641 -260 -253 -499 -321 -375 Commercial bank (net) 29 -61 -276 -44 27 71 14 -26 -88 267 125 -41 -356 -274 -355 0 131 -218 164 -384 365 811 313 76 -131 218 -164 384 -365 -811 -401 -76 -131 218 -164 384 -365 -811 -313 -76 li Inc1udes:a)Goodsprocuredinports; b)Repairson goods; c) Internalsales ofbondedcommodities. Excludes: Localsale 2/ Includes:a) Goodsprocuredinports; b) Repairson goods. Excludes:Freight& Insurancecharges 3/ Excludes:JDR grants 4/ Includes:JDR grants 5/ Excludes: Supplier'scredit Source: IMF 76 Belgium $327 million 77 FY96 FY97 FY98 FY99 FYOO FYOl FY02 FY03 FY04 Source: ExportPromotion Bureau 78 Table 5: Growth of Exports d RMG (us; 1illion) Grow '%) Total Woven Total Woven Exports RMG garments Knitwear Exports RMG garments