Embed Size (px)
Citation preview
This article was downloaded by: [Universidad de Sevilla]On: 13 November 2014, At: 01:31Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,UK
The Journal of DevelopmentStudiesPublication details, including instructions for authorsand subscription information:http://www.tandfonline.com/loi/fjds20
The end of textiles quotas: Acase study of the impact onBangladeshYongzheng Yang a & Montfort Mlachila aa International Monetary Fund , Washington, DC, USAPublished online: 04 May 2007.
To cite this article: Yongzheng Yang & Montfort Mlachila (2007) The end of textilesquotas: A case study of the impact on Bangladesh, The Journal of Development Studies,43:4, 675-699, DOI: 10.1080/00220380701259939
To link to this article: http://dx.doi.org/10.1080/00220380701259939
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all theinformation (the “Content”) contained in the publications on our platform.However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, orsuitability for any purpose of the Content. Any opinions and views expressedin this publication are the opinions and views of the authors, and are not theviews of or endorsed by Taylor & Francis. The accuracy of the Content shouldnot be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions,claims, proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly in connectionwith, in relation to or arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes.Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly
forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
The End of Textiles Quotas: A Case Studyof the Impact on Bangladesh
YONGZHENG YANG & MONTFORT MLACHILAInternational Monetary Fund, Washington DC, USA
Final version received August 2005
ABSTRACT This paper assesses the effects on the Bangladeshi economy of phasing out textileand clothing quotas by industrial countries. Bangladesh relies heavily on textile and clothingexports and is potentially very vulnerable to the abolition of the quotas. We used up-to-dateinformation on quota prices for Bangladesh to evaluate its competitiveness in a quota-free world,and subsequently incorporate the quota price information in the GTAP model to simulate theeffects of quota phase out on Bangladesh. We also examine in detail the supply constraints facingBangladesh’s textiles and clothing industries. Based on this analysis, we conclude that withoutaccelerated structural reforms Bangladesh is likely to face significant pressure on its balance ofpayments, output and employment in the aftermath of quota removal.
I. Introduction
In accordance with the WTO Agreement on Textiles and Clothing (ATC), allremaining textile and clothing (T&C) quotas maintained by industrial countriesunder the now defunct Multifiber Arrangement (MFA) were removed on 1 January2005. During the 10-year transition period 1995–2005, the remaining quotas werealso enlarged.1 Because these quotas were bilateral and the extent of theirrestrictiveness varied from country to country, their removal has altered thecompetitiveness of individual exporting countries. Countries that have faced morerestrictive quotas will see their competitive position improve after the quotas wereremoved, while those that have been less restricted by quotas may face difficultiesmaintaining their current market shares.2 The intensity of these shifts incompetitiveness will be amplified by the effective backloading of the quota phase-out under the ATC. Most of the restrictive quotas were removed only at the end ofthe transition period, turning what could have been a gradual adjustment processinto a major shock at the beginning of 2005.3
Bangladesh depends heavily on the exports of textiles and clothing, or ready-madegarments (RMG),4 and is potentially vulnerable to the large shock of the final stage
Correspondence Address: Yongzheng Yang, International Monetary Fund, 700 19th Street, N.W.
Washington, DC, 20431, USA. Email: [email protected]
Journal of Development Studies,Vol. 43, No. 4, 675–699, May 2007
ISSN 0022-0388 Print/1743-9140 Online/07/040675-25 ª 2007 Taylor & Francis
DOI: 10.1080/00220380701259939
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
of the quota phase-out. In 2002, these exports accounted for over 80 per cent of thecountry’s total merchandise exports – one of the highest shares among majorexporting countries (Figure 1). At the same time, Bangladesh depends on formerlyquota-restrained markets for about 94 per cent of its RMG exports, among thehighest ratios in the world. 5 Thus, the balance of payments consequence of a sharpdecline in RMG exports could be severe. The key question is whether Bangladesh’scompetitiveness will be weakened by the removal of the quotas and whether it will beable to maintain the rapid export growth that it achieved under the quota system.Falling exports could also have a significant impact on domestic production and
employment. The RMG industry has been the main source of growth in exports andformal employment in Bangladesh, although its direct contribution to GDP, atabout 5 per cent, is relatively small (Table 1).6 The industry plays a key role in
Figure 1. Proportion of T&C exports in total exports, 2002. Source: WTO and author’sestimates
Table 1. Bangladesh: growth of the RMG sector
Year
Exportvolume
(‘000 doz.pieces)
Export(US$ million)
Share intotal exports(per cent)
Employment(millions)
Number ofgarmentfactories
1985–86 4,763 131 16.0 0.2 5941990–91 30,567 867 50.5 0.4 8341995–96 72,005 2,547 65.6 1.3 2,3531999–2000 111,906 4,349 75.6 1.6 3,2002001–02 140,445 4,583 76.6 1.8 3,618
Source: Ahmed and Sattar (2003).
676 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
employment and in the provision of income to the poor, directly employing about1.8 million people, or about 40 per cent of manufacturing sector employment, 90 percent of whom are women (USITC 2004). The industry supports indirectly about 10–15 million people.7 Over the past 20 years, the number of manufacturing units hasgrown from 180 to over 3,600, 95 per cent of which are locally owned. The typicalfirm employs 200–1,200 workers, with an average of about 550–600 workers. Some90 per cent of the factories are located in and around the capital, Dhaka, and theport of Chittagong.
This paper evaluates the impact on the Bangladeshi economy of the phase-out oftextile and clothing quotas, with a particular focus on the medium-term effects on thebalance of payments (especially the trade account), GDP and employment. We usedup-to-date information on quota prices for Bangladesh to evaluate its competitive-ness in a quota-free world (Section II). We also examine in detail the supplyconstraints facing Bangladesh’s textiles and clothing industries (Section III). Wesubsequently incorporate the quota price information in the GTAP model tosimulate the effects of quota phase-out on Bangladesh (Section IV). Finally, in theconcluding section, we discuss policy implications of our analysis for Bangladesh.
II. Evaluating Competitiveness
Growth Record and Trade Preferences
Bangladesh’s RMG exports have grown rapidly over the past two decades, especiallyfollowing extensive trade and other economic reforms in the early 1990s. The valueof exports in US dollars increased more than sixfold during the period 1990–2002, orabout 16 per cent per year, considerably faster than the growth of the country’s othermerchandise exports (Figure 2).
Figure 2. Bangladesh: export performance, 1990–2003 (in millions of US dollars). Sources:Bangladeshi authorities and authors’ estimates
The End of Textiles Quotas 677
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Starting in 1998, however, RMG export performance slackened, with the annualgrowth rate averaging less than 5 per cent to 2002. For the first time in more than 20years, export values (in US dollars) actually declined in 2001–02. According toBangladeshi exporters, this is partly attributed to increasing competition in the worldmarket, especially from China and India, in a period of sluggish global demand. Theexporters also blame US preferential trade agreements for their export decline. Inparticular, they believe that the US Africa Growth and Opportunity Act (AGOA)has diverted US textiles imports away from Bangladesh to Africa.8
RMG exports recovered in the first half of 2003. Bangladeshi exporters believethat the rebound was mainly a result of a shift in demand in their favour due to theSARS epidemic that hit China and South East Asia in early 2003, duty- and quota-free access to the Canadian market beginning in January 2003, and strengtheningglobal demand. The recovery was also boosted by increased exports to the EU,where Bangladesh seems to begin to benefit from the Everything But Arms (EBA)Initiative. Under that Initiative, which entered into force in March 2001,Bangladesh, together with 49 other least developed countries (LDCs), benefits fromduty- and quota-free access for all products except arms (with phase-in periods forrice, sugar and bananas).On 1 January 2005, US quotas on exports from Bangladesh, along with those
from all other sources, were eliminated, but tariff restrictions will continue. In otherrestricted markets, both quotas and tariffs on Bangladeshi exports had alreadybeen removed before 1 January 2005.9 In contrast, most of Bangladesh’s majorcompetitors remained under quota restrictions in all major restricted markets until 1January 2005 (with the exception of Norway) and will continue to face tariffs.Quota-free access to the restricted markets (other than the US) gave Bangladesh adistinctive advantage over its competitors, and additional duty-free access (againexcept to the US market) further strengthened that advantage.10 The tariff preferencefor Bangladesh will remain even after quotas on all exporting countries wereterminated at the beginning of 2005.11 It should be noted, however, that rules oforigin have often limited the benefits of quota- and duty-free access. In the case ofthe EBA, while Bangladesh’s knit garments, which have high domestic value added(up to 80 per cent), can generally meet the requirement of 51 per cent domestic andregional valued added to be eligible for preferential access, its woven garments,which rely heavily on imported inputs, face a considerable constraint in meeting thisrequirement.12 Given that well over half of Bangladesh’s garments exports to the EUare woven products, this constraint is significant in determining the country’s overallexport performance in the EU market.13 It is reported that less than half ofBangladesh’s exports actually receives duty-free treatment under the initiative.14
Bangladesh is heavily dependent on the EU and US for its RMG exports. The twomarkets combined account for about 94 per cent of Bangladesh’s total RMG exports(Table 2). Exports to unrestricted markets are negligible. In comparison, most ofBangladesh’s main competitors ship a significant proportion of their T&C exports tothe unrestricted markets. China, for instance, ships over three quarters of its exportsto these markets, while for India the share is 40 per cent. In general, unrestrictedmarkets have been important for the more established exporters. Among the world’stop 25 clothing exporting countries, Bangladesh is probably the second mostdependent on the restricted markets after Macao SAR.15 These comparisons show
678 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Table
2.Bangladesh:directionoftradeforT&C
exports(inper
cent)
Woven
Garm
ents
KnitGaments
Total
1999/00
2000/01
2001/02
2002/03
1999/00
2000/01
2001/02
2002/03
1999/00
2000/01
2001/02
2002/03
Total
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
EuropeanUnion
41.8
41.8
44.5
47.7
69.6
70.1
69.8
73.0
49.9
50.5
52.6
56.2
United
States
54.2
54.3
47.2
46.6
25.4
24.9
24.9
21.2
45.8
45.2
40.1
38.0
Canada
2.1
2.1
2.0
2.9
2.8
2.4
2.4
2.9
2.3
2.2
2.1
2.9
Norw
ay
0.5
0.3
0.4
0.5
0.6
0.4
0.6
0.7
0.5
0.4
0.5
0.6
Switzerland
0.3
0.3
0.4
0.5
0.6
0.7
0.6
0.6
0.4
0.4
0.5
0.6
South
Korea
0.1
0.1
0.1
0.1
0.0
0.0
0.0
0.0
0.1
0.1
0.1
0.1
Japan
0.2
0.2
0.4
0.4
0.3
0.3
0.3
0.2
0.2
0.3
0.4
0.3
Australia
0.1
0.1
0.1
0.1
0.0
0.0
0.1
0.1
0.1
0.1
0.1
0.1
Other
0.8
0.9
5.0
1.3
0.8
1.2
1.3
1.1
0.8
1.0
3.8
1.2
Mem
orandum
item
EU
andUS
96.0
96.0
91.7
94.2
94.9
95.0
94.7
94.2
95.7
95.7
92.7
94.2
Source:
BangladeshBankandauthors’calculations.
The End of Textiles Quotas 679
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
that Bangladesh has yet to demonstrate a capacity to penetrate unrestricted markets.They also imply that Bangladesh may benefit little from price increases in theunrestricted markets when demand in the restricted markets increases with theelimination of quotas.16
How Bangladesh Will Be Affected by Quota Removal
In the US market, Bangladesh’s exports are more concentrated in formerly quota-restrained products than most of its competitors. Bangladesh faced quotas in 30categories of products. Although this was low compared to 90 categories for China, itwas similar to most other exporting countries. In value terms, however, Bangladeshhad higher quota coverage than all but Hong Kong and Sri Lanka (Table 3). This canindicate either more comprehensive restrictions on Bangladeshi exports or moregenerous access to quotas. In either case, the impact of the quota removal onBangladesh (positive in the former case and negative in the latter case) would tend tobe larger than on other countries. However, without knowing the restrictiveness ofthe quotas on Bangladesh’s exports versus those on its competitors’ it is difficult toassess Bangladesh’s competitiveness based on quota coverage. Bangladesh’s highquota utilisation rate suggests that it faced binding quotas, but this does not implythat it faced more restrictive quotas than its competitors. A useful indication of therestrictiveness of a quota is the price it commanded in the market. In principle, thehigher the price, the more restrictive the quota.Quota price data indicate that Bangladesh’s quotas probably fell in the middle of
the restrictiveness scale. Based on 2002–03 quota price data reported by the WorldBank (Table 4), Bangladesh appears to be the second most restricted Asian country
Table 3. United States: textile and clothing quotas, 2001–02 (in millions of US dollars, and percent)
Numberof quotacategories
Averagequotafill rate
Importsunder
Quotas (a)Total Imports
(b)Percent(a)/(b)
2001 2002 2001 2002 2001 2002
Bangladesh 30 83 1,453 1,396 2,235 2,017 65 69Cambodia 23 24 548 639 953 1,062 57 60China 90 76 4,669 5,315 9,629 11,476 48 46Egypt 19 10 154 141 515 493 30 29Hong Kong 64 55 3,848 3,809 4,461 4,081 86 93India 30 68 1,497 1,714 2,912 3,294 51 52Indonesia 34 33 1,109 1,045 2,586 2,363 43 44Pakistan 36 31 1,066 1,047 1,958 2,010 54 52Philippines 42 32 1,437 1,460 2,274 2,060 63 71Sri Lanka 38 26 1,132 1,065 1,725 1,552 66 69Thailand 59 56 1,468 1,470 2,534 2,299 58 64Turkey 28 23 850 990 1,472 1,702 58 58
Source: United States Department of Commerce.Note: Quota categories aggregated at 3—digit classification.
680 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
after China, which is widely regarded as the most restricted exporter in the USmarket (USITC, 2004). However, data provided by Bangladeshi industry sourcesindicate that quota prices in Bangladesh may have fallen since 2002. While the WorldBank estimate for 2002 is 20 per cent of the fob price (net of the quota price), whichis substantially lower that that for China (36 per cent), similar to that for India(20 per cent), and considerably higher than that for Pakistan (10 per cent), the latestdata suggest that Bangladesh’s average quota price had fallen to about 7.6 per centin late 2003 and early 2004 (Table 5). It should be noted, however, that suchestimates are inherently volatile given the nature of quota restrictions. Therestrictiveness of a quota as measured by its price varies with demand and supplyconditions over time.
Comparisons with competitors at the commodity level suggest that Bangladeshiexporters are likely to be subject to intense competition in the US market now thatthe remaining quotas have been eliminated. For every Bangladeshi product restrictedby quota its Chinese counterpart was also restricted (Table 5). The Finger-Kreininsimilarity index indicates that the two countries’ exports to the US overlapped by asmuch 72 per cent (Table 6).17 Nine out of 10 of Bangladesh’s top exports coincidedwith China’s top 10. While quota utilisation rates were similar for the two countries,quota prices were much higher in China than in Bangladesh across all quotas.China’s average quota price (using China’s trade weights) as a percentage of the fobprice (net of quota rent) was 40 per cent for products for which Bangladesh was alsorestricted. Using Bangladesh’s trade weights, China’s average quota price would be49 per cent, over six times the average Bangladeshi price (7.6 per cent). The same wastrue for quota utilisation. China’s average utilisation rate for products for whichBangladesh was also restricted was 16 percentage points higher than its overallaverage (72 per cent). Bangladesh’s export similarity with India and Pakistan is lowerthan with China, but still significant. The lower similarity results from India andPakistan’s greater specialisation in textile products, in contrast to Bangladesh’sheavy concentration in garments.
Table 4. Estimated export tax equivalents of quotas in key supplying regions, 2002–03 (percent of fob prices net of quota rents)
Textiles Clothing
USA EU USA EU
Bangladesh1 0.0 0.0 20.4 0.0India 3.0 1.0 20.0 20.0Pakistan1 9.8 9.4 10.3 9.2China1 20.0 1.0 36.0 54.0Hong Kong SAR1 0.0 2.1 2.3 12.3Sri Lanka 0.0 1.0 7.0 0.0Other East Asia2 0.0 1.0 7.0 3.0Newly industrializing economies3 0.0 1.0 2.5 0.3
Source: World Bank (2004).Notes: (1) Denotes an estimate based on quota price information. Other estimates areinterpolated from quota utilization data. (2) Based on Indonesia, Philippines, Thailand. (3)Republic of Korea and Taiwan Province of China.
The End of Textiles Quotas 681
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Table
5.BangladeshandChina’sexportsto
theUnited
States:values,quota
utilizationrates,andquota
prices,2003*
Bangladesh
China
Export
value
Quota
fillrate
Quota
price
Export
value
Quota
fillrate
Quota
price
Cat.
Product
(US$million)
(per
cent)
(per
cent)**
(US$million)
(per
cent)
(per
cent)**
237
Playsuits,etc
10
85.6
n.a.
80
63.2
5.7
331
Gloves
158.3
n.a.
35
90.0
7.3
334
Coats,non–suit,M&B
20
93.3
13.6
66
89.7
26.9
335
Coats,W&G
24
86.8
1.5
79
94.1
28.0
336/636
Dresses
31
79.0
1.8
262
81.5
10.8
338/339
Knitshirt&
blouses
115
91.8
2.8
206
94.3
157.8
340/640
Shirts,notknit,M&B
226
94.7
0.6
155
94.2
74.2
341
Shirts
&blouses,notknit,W&G
94
78.4
0.3
80
93.1
37.3
342/642
Skirts
38
83.9
5.0
85
95.1
42.1
347/348
Trousers,etc
263
96.5
10.0
315
88.4
44.9
351/651
Nightw
ear
49
93.5
2.5
111
87.6
41.3
352/652
Underwear
118
90.2
0.3
104
85.8
35.9
363
Terry
&other
piletowels
18
77.8
n.a.
69
92.4
n.a.
369–S*
Shoptowels
25
86.1
n.a.
n.a.
n.a.
n.a.
634
Coats,non–suit,M&B
80
94.4
2.5
180
81.0
15.9
635
Coats,W&G
55
96.2
2.5
171
88.0
24.2
638/639
Knitshirts
&blouses
68
91.9
0.6
204
92.4
43.1
641
Shirts
&blouses,notknit,W&G
28
80.0
0.6
108
89.2
24.9
645/646
Sweaters
17
84.9
1.8
78
85.9
35.4
647/648
Trousers,etc
115
94.6
3.0
285
88.4
20.8
847
Trousers,breeches
&shorts
16
76.7
n.a.
n.a.
n.a.
n.a.
Total/average***
1,412
86.4
7.6
2,671
88.1
40.3
TotalT&C
exportsto
theUS
1,990
8,744
Source:
Export
values
andquota
fillratesare
from
theU.S.Office
ofTextilesandApparel,quota
pricesforBangladeshare
from
communications
withBangladeshitraders,andquota
pricesforChinaare
from
ChinaQuota.com
athttp://w
ww.chinaquota.com/EN/index.asp.
Notes:*Quota
pricespertain
tolate
2003andearly2004forBangladeshandthewhole2003forChina;**Per
centofthefobprice
net
ofquota
rents;
***Averages
are
weightedbytradevalues.
682 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Bangladesh’s relatively low quota prices resulted partly from the generous quotaallocation it received in the US market.18 The growth rates of quotas on Bangladeshiexports were higher than those on exports from virtually all of its main competitors(Table 7).19 By 2002, Bangladesh was entitled to an average annual growth rate of12.9 per cent for its base quotas, 60 per cent higher than the average for its maincompetitors.20 In fact, for a number of products Bangladesh had larger quotas thanChina. Weighted by Chinese unit values (CIF prices), which on average were doubleBangladeshi unit values, Bangladesh’s aggregate quota for its restricted productswas more than 20 per cent larger than China’s. In other words, if Bangladesh hadachieved China’s level of efficiency and product quality, its quotas in the US wouldhave been worth US$3.2 billion. With the current average quota utilisation rate, itcould have doubled its actual value of exports to the US without any increase inquota volumes.
In the EU market, the fact that Bangladesh did not face any quotas there meansthat the shock to its exports could be even larger after quotas on other countries were
Table 6. Export similarity between Bangladesh and its major competitors, 2002 (in per cent)
Competitor USA EU
China 71.5 22.0India 57.1 39.1Pakistan 34.8 67.6
Source: Authors’ estimation based on data from US Department of Commerce and theEuropean Commission.
Table 7. Growth rates of US base quotas1 for Bangladesh and its main competitors (per centper year)
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Bangladesh 8.1 8.1 8.1 10.2 10.2 10.2 10.2 12.9 12.9 12.9China – – 1.7 0.1 1.7 1.7 1.7 0.4 2.4 2.0India 6.7 6.7 6.7 8.4 8.4 8.4 8.4 10.6 10.6 10.6Indonesia 6.2 6.2 5.5 7.7 7.7 7.7 7.7 8.3 9.8 9.8Pakistan 7.6 11.9 7.6 9.2 9.4 9.5 9.5 14.9 12.0 12.0Sri Lanka 6.1 6.2 6.2 7.7 7.7 7.7 7.7 6.3 9.7 9.7Thailand 6.1 6.1 6.1 6.0 7.6 7.6 7.6 7.6 9.6 9.6Vietnam 2/ – – – – – – – – – 6.0Memo item:Simple average
excl. Bangladesh6.5 7.4 5.6 6.5 7.1 7.1 7.1 8.0 9.0 8.5
Sources: US Office of Textile and Apparel, and authors’ calculations.Notes: (1) Note that these are different from ‘‘adjusted’’ limits available from the archivedDecember 31 Customs Quota Status reports. The base quotas are often adjusted by ‘‘swing’’(increasing a quota by up to 7 per cent, by borrowing from another category), ‘special shift’(similar to swing, but in addition to swing), ‘carryforward’ (borrowing up to 6 percent fromthe next year’s quota) and ‘carryover’ (using unused quota, up to 11 percent of the receivingyear’s limit, from the previous year). (2) Prior to May 2003, Vietnam did not have any quotas.
The End of Textiles Quotas 683
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
removed. Unlike in the US market, where Bangladesh had a 7.6 per cent quota rentto cushion price declines after quotas were removed, any price falls in the EU marketwould directly cut into profits and hence exert pressure on exports. Estimates ofETEs suggest that the prices of exports from Bangladesh’s competitors couldindeed fall considerably after the quotas were removed in the EU market. In fact,Chinese clothing exporters seem to have faced higher average ETEs in the EU thanin the US, while Indian and Pakistani exporters faced similar levels of ETEs in bothmarkets (Table 4). If the price elasticity of demand for Bangladeshi exports is alsothe same in the EU as in the US market, the downward pressure on Bangladeshiexports in the EU market would be larger than in the US market following the quotaremoval.21
The degree of substitution between exports from Bangladesh and its maincompetitors is potentially high in the EUmarket. At first sight, this does not appear tobe the case based on estimates of export similarity in the EU market. At the quotalevel, the overlap between exports from Bangladesh and China is low, and that withIndia and Pakistan is considerably higher (Table 7).22 Export similarity betweenBangladesh and China is low because China’s exports to the EU are very diversified,while Bangladesh is specialised in several major clothing categories. However, just fiveof these categories accounted for 86 per cent of Bangladesh’s total RMG exports tothe EU in 2002, while they made up only 13 per cent of China’s exports. In valueterms, Bangladesh exported nearly twice as much of these products as China did.Remarkably, according to World Bank (2004) estimates, China’s exports in thesecategories faced very high ETEs.23 Pakistan also had higher-than-average ETEs forthese categories, in addition to a higher share of these products in its total exports tothe EU, which contributes significantly to the higher export similarity between thetwo countries.24 Thus, despite the apparent low export similarity between Bangladeshand China, competition between the two countries is likely to be intense.25
Performance During the Transition
The strong competition between Bangladesh and other exporters seems to have beenborne out by the recent developments in exports for which quotas were removed atthe beginning of 2002 during the Phase III quota integration (2002–04) under theATC. Export performance in these quota categories thus provides initial evidence ofcompetitiveness after 2004. From 2001 to 2003, while Chinese exports surged,Bangladeshi exports of products that fall under Phase III integration declined by 46per cent (in value terms) in the EU market and 41 per cent in the US market(Figure 3). Among the seven categories of products for which quotas were removed inthe US market in 2002, Bangladesh suffered export losses in all but one category.Many other countries also suffered export declines, but Bangladesh’s loss in the EUmarket might seem surprising given that it had quota-free and duty-free access underthe EBA.
III. Domestic Supply Constraints
In addition to generous quota access to major export markets, Bangladesh’s keycompetitive advantage is its low wages. Unit labour costs are, however, higher than
684 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Figure 3. Changes in the value of T&C products under Phase III of quota integration(percentage change January–September 2003 vs. January–September 2001). Source: EU, US
International Trade Commission, and authors’ estimates
The End of Textiles Quotas 685
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
in India and China, because of lower labour productivity, as measured by valueadded per worker (Table 8). This lower productivity is a result of a number of supplyconstraints that need to be overcome if Bangladesh is to improve its competitiveness(Dowlah, 1999).
Structural Weaknesses
Structural rigidities have made it difficult for Bangladesh to fully exploit its low wageadvantage. In a recent World Bank-led study (Bangladesh Enterprise Institute andWorld Bank, 2003), the following are identified as key structural constraints toinvestment in Bangladesh:
. Defective and insufficient infrastructure poses some of the most severe obstaclesto companies in Bangladesh.
. Corruption is pervasive and costly. It often manifests itself in excessiveregulation, leading to extortion and bribery. Companies rank it as a severeobstacle to business.
. High levels of nonperforming loans reduce the capacity of banks to lend atreasonable interest rates, especially to small and medium-sized enterprises.
While weak infrastructure is an impediment to activities across all industries, itoften imposes a critical constraint on export-oriented industries, such as textiles andclothing, where competitive prices, consistent quality and reliable, fast delivery arevital for export success. Electricity supply continues to be a bottleneck despiteconsiderable FDI received in the power sector over the past decade. About 70 percent of companies rely on back-up generators that supply electricity at a costtypically 50 per cent higher than the price of power from the public grid. Bangladeshhas expensive and often inaccessible telecommunications networks (especially for
Table 8. Selected characteristics of the garments sector
Valueaddedper
employee
Wagesand
salariesper
employee
Unitlabourcost
Materialsand
utilitiesCosts oflabour
Operatingsurplus
Country Latest year (in US dollars) (in per cent of output)
Bangladesh 1997 900 400 0.44 75.4 9.7 15.0China 2001 5,000 1,600 0.32 74.8 13.7 11.5El Salvador 1998 5,100 2,500 0.49 30.7 33.5 35.7Hong Kong
SAR1999 27,600 14,800 0.54 71.1 15.5 13.4
India 1998 2,600 700 0.27 77.8 6.3 15.8Indonesia 1999 2,500 600 0.24 63.4 8.5 28.1Morocco 1998 4,000 2,500 0.63 55.9 27.9 16.2Sri Lanka 1998 2,500 700 0.28 53.4 13.6 33.0
Source: European Union (2003); China statistical yearbook 2002; World Bank (2004).
686 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
overseas connections). The country has significantly lower intensity of telephonelines and higher international rates than its major export competitors. AlthoughBangladesh has a reasonably developed nationwide road network, there isconsiderable congestion on the roads. Bangladesh’s main export gateway, the portof Chittagong, is very slow in handling containers due to lack of cranes. Thedevelopment of a privately owned container terminal at Chittagong has been slow asa result of labour disputes. Increasingly, exporters have resorted to air freight toavoid losing orders, thereby further squeezing profit margins. In an era in whichexport success is increasingly dependent on rapid response and quick turnaround oforders, Bangladesh’s poor infrastructure places its exporters at a distinctivedisadvantage against their competitors (Cookson, 2003b; Spinanger and Wogart,2001).
In the post-ATC textile and clothing market, infrastructure costs will become anincreasingly important determinant of FDI inflows and import sourcing.26 Asurvey of major investors conducted in Hong Kong in 2000 showed that apartfrom quotas, political stability, policy predictability, and good transportinfrastructure were the most important factors in influencing FDI in the RMGsector (Spinanger and Wogart, 2001). Interestingly, lower wages, if not associatedwith reasonable productivity, were not a very important factor. A follow-up survey(Spinanger and Verma, 2003) conducted in 2003 showed that transportinfrastructure had become even more important while the availability of quotaswas actually no longer a factor for investment decisions as the final stage of quotaintegration drew closer. The importance of transport infrastructure in determiningimport sourcing is highlighted by a recent study that finds that each additional dayin transit is equivalent to an extra 0.8 percentage point increase in applied advalorem tariffs (Hummels, 2001).
The perception of widespread corruption further reduces the attractiveness ofBangladesh as an FDI destination and source of imports. In 2003 Bangladesh wasranked last out of 133 countries in the Transparency International CorruptionPerception Index (CPI). Furthermore, excessive and capricious bureaucratic controlswith large discretion in their implementation, especially licensing requirements andcomplicated customs procedures, are viewed as a major impediment to FDI andtrade. Finally, weak law and order, and especially widespread extortion, hinderforeign investment.
Policy-Induced Rigidities
In addition to the above structural weaknesses, there are a number of policy-inducedrigidities that have reduced the competitiveness of the textile and clothing sector.These include: (1) restrictions on FDI in the sector; (2) a requirement to have back-to-back letters of credit (LCs) before imports can be approved; (3) a requirement toreserve 40 per cent of export cargo for domestic vessels; and (4) an inefficient quotaallocation system, although this is no longer an issue with the removal of thequotas.27
Except in the EPZ, FDI in the RMG sector is severely restricted to protect localproducers. All new FDI in the sector has to be vetted by the Board of Investment(BoI) in consultation with the RMG Manufacturers’ Association, which feared
The End of Textiles Quotas 687
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
competition for quota allocation. Although the latter has no formal veto power, inpractice the BoI has followed the association’s recommendations. In general, themanufacturers only recommend approval of FDI in sub-sectors where Bangladesh isdeemed not competitive.28
In restricting the role of FDI in the broader textile and garment industries,Bangladesh has foregone a number of FDI-related benefits. Foreign investors oftenbring superior technology and managerial skills. FDI also helps local firms join theglobal value chains, which are dominated by multinational companies based inindustrial countries (Gereffi, 1999). Such value chains are often critical sources ofproduct information and channels of marketing. Greater integration with the globalvalue chains has enabled Bangladesh’s competitors to move faster to higher qualitymerchandise, usually under brand names, where the profit margins are better. Asnoted earlier, Bangladesh’s average unit prices for major product categories areconsiderably lower than for corresponding Chinese products. Of course, FDIrestrictions only partly explain Bangladesh’s low export prices. Other factors includeinadequate labour training, outdated equipment, poor infrastructure, and therelatively large quotas that gave Bangladeshi exporters fewer incentives for productupgrading.Current global value chains are characterised by consolidation of sources of
supply and increased involvement of retailers in product sourcing, quality control,and the setting of labour and environmental standards (Gherzi Textil and others,2002). As a result, the role of pure buying houses that Bangladesh relies on isdeclining.29 Consolidation in Bangladesh is already taking place among the largermanufacturers who are pulling together factories in one location with larger andbetter facilities and equipment. Some large operators are relocating to areas ofBangladesh away from large cities, where labour costs are lower. Smaller operatorswho are unable to consolidate their operations are most likely to suffer, especiallygiven Bangladeshi business practice whereby large operators rarely buy outstruggling small factories.The requirement of back-to-back LCs significantly increases the lead times for
exporters. With a few exceptions, imports of raw materials for use in the RMG canonly be approved if there are back-to-back LCs.30 In other words, approval forimports is granted if an importer can prove that he already has an export orderbacked by an LC. The justification for this rule is to prevent non-exporters frombenefiting from the duty-free access to inputs that is available to exporters. Inpractice, this rule benefits greatly the domestic producers of textiles used in theproduction of garments because of the effective lengthening of delivery time forimported inputs. Not surprisingly, the main support for this system of back-to-backLCs comes from textiles manufacturers.Under the Flag Protection Ordinance (1982), exporters are required to set apart 40
percent of export cargo for domestic vessels. The law has not been effectivelyenforced as it is difficult in practice to split individual shipments 40:60 per cent.Exporters usually get around the requirement by seeking waivers from the office ofthe Director General of Shipping. While the law’s effectiveness is debatable, itimposes considerable costs on exporters in terms of lost time and potentialcorruption. The recent ban on the use of land ports for imported RMG inputs fromIndia is also likely to further increase export costs.
688 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Like in many other countries, quota allocation in Bangladesh was essentiallybased on past performance.31 The system was backward-looking and failed toencourage competition. It tended to favour large and well-established firms at theexpense of newer, potentially more innovative firms. Despite the demise of thesystem with the removal of the quotas on January 1, 2005, it may take time toeliminate the inefficiencies created by the system. To realize the efficiency gains,Bangladesh needs to ensure competition in the textile and garment industries andreduce rent-seeking activities that may continue to exist due to other governmentregulations.
IV. Estimating the Effects of Quota Removal
Several studies have attempted to assess the impact of quota removal onBangladesh’s economy, especially on RMG exports (Table 9). It is difficult to drawreliable conclusions from these studies: (i) some are based on past experience; (ii)some are not quantified; (iii) some are really based on conjecture; and (iv) thecoverage of policy change varies. This notwithstanding, most studies agree that theimpact is likely to be negative if the Bangladeshi government and industry do little toaddress key impediments to export expansion. For example, on the basis ofexperiences whereby Sweden and Canada removed quotas in 1991 and 1998,respectively, leading to a loss of market share for Bangladesh mainly to China,Spinanger and Wogart (2001) conclude that there is a high probability that a loss inmarket share will occur after 2004.32 Cookson (2003b) estimates that about 50 percent of the US market and 35 per cent of the EU market could be lost tocompetition, leading to an overall loss of 35 per cent of RMG exports.33 Using thestandard GTAP model, Herok and van Tongeren (2002) estimate that Bangladeshwould be able to maintain its current level of export volumes despite a fall in theexport price. 34 Introducing economies of scale to the same model, Lips and others(2003) estimate that clothing output would fall by as much 20 per cent and exportsby 6 per cent.35
A general equilibrium approach is essential to capture the economy-wide effects ofquota removal by taking into account changes in economic aggregates (for example,income and investment) and sectoral linkages. In this section, we also use the GTAPmodel to estimate the impact of the quota phase-out on the Bangladeshi economy. Inaddition to a key refinement of quota price data, we explore a different labourmarket closure (which enables us to estimate employment impact) and otherpotential market outcomes (see the various scenarios below. Also see AppendixTable A1 and A2 for commodity and regional aggregations). To more accuratelyreflect the current extent of quota restrictions, data on quota premiums are updatedbased on the latest estimates for Bangladesh (Table 5). Such estimates play a criticalrole as they indicate the relative restrictiveness of quotas across exporting countries.Most of the existing studies rely on estimates for the mid-late 1990s. As noted earlier,Bangladesh’s quota prices may have declined prior to the quota removal, perhapsindicating weakening competitiveness.
The simulations focus on the static, medium-term effects of quota removal. Forthis reason, the database of the model is updated to 2007 through a projectionexercise, which involves augmenting GDP, population and factor (land, labour,
The End of Textiles Quotas 689
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
capital and natural resources) endowments with productivity accounting for anyslack in GDP growth.36 GDP and employment projections are based on IMF WorldEconomic Outlook (September 2003), while population projections are based on theWorld Bank World Development Indicators (2002). Capital accumulation projections
Table 9. Bangladesh: a summary of findings on the effects of MFA quota removal
Study Estimated impact Methodology Notes
Cookson (2003a) 35 per cent decline inexport values; 50/20per cent of U.S./EUmarket lost
Interviews with mainexporters andauthor’s conjecture
No rigorous analysis
Gherzi Textil andothers (2002)
Presuming a negativeimpact but noestimates areavailable
No quantification ofthe impact
Focuses on policyrecommendationsto the authorities
Spinanger andVerma (2003)
GDP: 70.14 per centOverall exports: 70.1
per centTextile exports: 15.5
per centClothing exports:
77.9 per cent
Simulations ofcombined effects ofquota eliminationand China’s WTOaccession using theGTAP model
Numbers bench–marked to a 1997baseline; fixedemployment level
Spinanger andWogart (2000)
Bangladesh’s share ofSwedish marketdeclined from 16 to3 per cent withinone year after the1990 elimination ofquotas
Ex post estimation.No isolation of theeffect of quotaremoval
Suggestive resultsfor effects of quotaremoval in 2005
Herok andTongeren(2002)
RMG export volume:unchanged
RMG export price:77 per cent
Share of world exportvolume: 2 per cent
Simulation ofcombined effects ofquota elimination,China WTOaccession, EUEasternenlargement andEU–Turkeypreferentialagreement
Numbers bench–marked to a 1997baseline; labourmarket closureunknown; resultssuggest a declinein world RMGexports
Lips, Tabeau andTongeren(2003)
Textile output: 71per cent
Clothing output:720per cent
Total export value:76.4 per cent
Total import value:76.2 per centWelfare loss (EV):
US$441 million in1997 prices
Simulation ofcombined effects ofquota elimination,China WTOaccession, EUEasternenlargement andEU–Turkeypreferentialagreement
Numbers bench–marked to a 1997baseline; presenceof economies ofscale; labourmarket closureunknown
Source: Authors’ compilation from the studies listed.
690 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
are guided by projected GDP growth and historical data provided in Hertel andothers (1996). Changes in arable land are based on Anderson and others (1996). Fornatural resources, constant prices are assumed over time and the level of resource useis determined endogenously.
In simulating the impact of the quota phase-out, quotas on exports from all otherdeveloping countries are also removed together with those on Bangladeshi exports(which face restrictions only in the US market, as noted earlier). No other policychanges are introduced. The removal of Canadian and EU quotas on Bangladeshiexports is incorporated in the baseline projections. The simulation does not takeaccount of any dynamic or non-price effects of the quota phase-out, such as impro-vements in product quality and transport facilities. Whether an exporting countryexperiences an export expansion or contraction after quota removal depends primarilyonwhether its production cost (net of quota premiums) is lower or higher than itsmaincompetitors’. 37 All results are reported as deviations from the 2007 baseline.
A number of scenarios, based on different assumptions on elasticities of substitutionand factor markets, are examined. It is assumed in most of the simulations that inexporting developing countries real wages remain constant, while employmentresponds to changes in demand.38 Labour and capital are assumed to be perfectlymobile across industries, but completely immobile internationally. Domestic invest-ment is determined by the expected rate of return, which is equalised (net of riskpremium) across countries through international movement of savings in search forhigher returns. Saving is a linear function of national income. Land is confined to theuse in agriculture, while natural resource use is associated with only mining activities.
Simulation results confirm the consensus that Bangladesh is likely to be adverselyaffected by the phase-out of textile and clothing quotas (Table 10).39 Under the firstscenario, in which standard GTAP elasticities are applied and real wages areassumed to be fixed, clothing exports fall substantially, while textile exports remainessentially unchanged.40 However, because of the great weight of clothing in totalexports, overall exports fall considerably. Bangladesh’s overall terms of tradedeteriorate. The extent of the impact on clothing exports is not surprising given theirheavy concentration in the restricted markets.41 Overall imports also fall, largely as aresult of declines in textile imports. On balance, the trade account deteriorates by 0.4per cent of GDP. Despite the relatively weak backward linkages of the garmentsindustry with the domestic textile industry and the rest of the economy, the effects ofquota removal on GDP and employment are large – and perhaps larger than thecurrent share of textile and clothing in GDP would suggest. GDP contracts by 1.3per cent, while employment declines by 2.1 per cent.
The simulation results are sensitive to the elasticities of substitution betweenproducts from different countries of origin. For this reason, it is important toundertake sensitivity analysis to gauge the range of plausible results. Intuitively, thegreater the substitutability between Bangladeshi and its competitors’ products, thelarger is the impact on Bangladesh’s exports when quotas are removed. Our earlieranalysis of product similarity between Bangladesh and its main competitors suggeststhat sensitivity analysis of higher elasticities is particularly important. As Scenarios 2and 3 in Table 10 show, lower elasticities (half the values of the central elasticities)would significantly reduce the impact on Bangladesh, while higher elasticities(double the values of the central elasticities) would imply a more dramatic impact on
The End of Textiles Quotas 691
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Table
10.Bangladesh:eff
ects
oftextile
andclothingquota
removal1(percentagedeviationfrom
thebaseline,
unless
otherwiseindicated)
Scenario1
Scenario2
Scenario3
Scenario4
Scenario5
Scenario6
Central
elasticities2
Lower
elasticities3
Higher
elasticities4
Constant
employment5
Lower
investm
ent6
China
restrictions7
GDP
71.3
70.8
72.2
70.3
72.0
71.0
Employment
72.1
71.5
73.2
0.0
73.5
71.7
Tradebalance
Inbillions1997US$
70.3
70.2
70.6
0.3
0.8
70.3
Inper
centofGDP
70.4
70.3
70.7
0.3
1.0
70.3
Termsoftrade
72.4
72.0
72.9
73.2
73.5
72.0
Totalexports
78.4
73.5
720.3
73.0
70.7
76.8
Clothing
711.8
74.9
729.0
76.2
73.9
79.5
Textiles
0.1
70.1
1.5
4.5
5.9
0.1
Totalim
ports
76.8
73.3
715.3
76.7
78.2
75.6
Clothing
79.0
74.1
720.2
711.4
712.3
77.7
Textiles
710.6
74.6
724.6
78.4
77.2
78.8
Clothingoutput
711.3
74.7
727.7
75.9
73.7
79.1
Textile
output
73.7
71.6
79.2
1.2
2.8
72.9
Source:
Sim
ulationswiththeGTAPmodel,asdescribed
inthetext.
Notes:
(1)Thepercentagechangenumbersare
notchanges
inthegrowth
ratesofthevariables.
Theseare
changes
inthelevelsofthevariables
benchmarked
againstthebaselinewhichassumes
textileandclothingquotaswould
continueto
existin
2007.(2)Nominalwages
are
fullyindexed
totheCPIwithmedium
levelsofelasticities.
(3)Halfofthecentralelasticities.
(4)Double
ofthecentralelasticities.
(5)Wages
are
fullyflexible
tomaintain
currentem
ployment.(6)Asaresultofonepercentagepointrise
intherisk
premium
forinvestm
entin
Bangladesh.(7)China’s
T&C
exportsto
theU.S.andEU
increase
only
byhalfofthose
under
scenario(1).
692 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Bangladesh. Nevertheless, within this wide range of elasticities, the direction of theimpact remains unchanged, and the qualitative results continue to hold.42
Factor market assumptions are critical in determining the impact of the quotaphase-out on macroeconomic aggregates. If instead of a fixed real wage outcome,Bangladesh’s RMG industries decide to maintain their current level of employment,the impact would be smaller. However, workers throughout the country would haveto accept a 1.2 per cent reduction in real wages (Scenario 4). The balance of trade infact improves slightly because of a considerable real exchange rate depreciation, asindicated by further deterioration of the terms of trade. Another importantassumption is how investment will be affected by quota removal. If investors believethat Bangladesh will be adversely affected by quota removal, there could be ademand for a higher expected rate of return on investment in Bangladesh. Sucha possibility is illustrated in Scenario 5, where it is assumed that investors perceive arise in the risk premium of 1 percentage point in Bangladesh. This would exacerbateGDP and employment contractions, but reduce the impact on the trade balance byfurther reducing investment.43
Given the recent pressure in the US and EU to reimpose quotas on Chinese textileand clothing exports after 2004, a simulation of restrained Chinese exports is alsocarried out (Scenario 6).44 It is assumed that as a result of a newly negotiatedarrangement, Chinese textile and clothing exports would increase by only half whatthey would if quotas were completely phased out and no new restrictions wereintroduced. The results indicate that the adverse impact on Bangladesh’s GDP,employment and exports would be about 20 per cent less than under the firstscenario, while the impact on the balance of payments declines marginally. Thedampened expansion of China’s exports is partially offset by increases in exportsfrom Bangladesh’s other competitors, such as India and ASEAN.
An increase in productivity would help offset the adverse effects of the quotaphase-out. Simulations indicate that to maintain the baseline level GDP, Bangladeshwould need to increase its total input productivity in the textile and clothing sector(relative to its competitors) by 4–5 percent (cumulatively) in 2007. To ensure baselinelevel employment, the sector needs to achieve a 5–6 per cent increase in productivity.Such productivity improvements, though not particularly large, would alsosubstantially reduce the potential deterioration of the trade balance. This under-scores the importance of overcoming supply constraints (e.g., weak infrastructure)analysed in the previous section.
V. Concluding Remarks
Wehave assessed from several angles the vulnerabilities of theBangladeshi economy tothe removal of textile and clothing quotas. Unlike most of the existing studies, whichtend to rely on quota price information for the 1990s, we used detailed, up-to-date dataon Bangladesh’s quota prices and utilisation, market and product distribution of itsexports, and export performance during the third phase of quota removal. Wesubsequently incorporated some of these data into the GTAP database for oursimulation exercise, with a series of sensitivity analyses to explore a plausible range ofoutcomes for the Bangladeshi economy. Furthermore, we examined in some detail thesupply constraints facing Bangladeshi garment exporters, pointing to policy direction
The End of Textiles Quotas 693
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
that the Bangladeshi government can take to increase the country’s competitiveness inthe post-ATC world market. Our main findings can be summarised as follows.Three factors seem to have contributed to Bangladesh’s impressive export
performance in the textile and clothing sector in the 1990s: low wages, initial FDIinflows, and generous quotas in the restrictedmarkets relative to its main competitors.Low wages and generous quotas initially attracted FDI and enabled rapid exportgrowth, but Bangladesh has not been able to capitalise on this by continuouslyupgrading and diversifying its RMG exports. In fact, large rents generated by quotasand government assistance may have weakened incentives to improve productivity.With the phase-out of remaining quotas on 1 January 2005, Bangladeshi exportersare facing increasing competition. The export slowdown since 1998 and evidencefrom the third phase of quota removal indicate that Bangladeshi exporters are likelyto find it difficult to maintain their market shares in the US and the EU after 2004.This relatively weak competitiveness makes the Bangladeshi economy highly
vulnerable to the final stage of the quota phase-out. Simulation results indicate thatBangladesh’s exports could fall substantially in the wake of quota removal and thatits balance of payments position could be weakened considerably. While the RMGsector’s contribution to GDP is relatively small, the impact of quota removal couldbe amplified through labour market rigidities as well as indirect effects throughbackward and forward linkages to the rest of the economy. The resulting pressureson domestic production and employment could also be severe.There is considerable uncertainty over the ultimate impact of the quota removal
on the Bangladeshi economy. If, for example, investors perceive a major weaknessin Bangladesh’s competitiveness after 2004, investment could contract further,exacerbating the negative impact on employment and output. There is also consi-derable uncertainty over the extent of labour market flexibility in Bangladesh andhow easily its products can be replaced by those from its competitors. Quotas couldbe reintroduced on Chinese exports to the advantage of Bangladeshi exports. ButBangladesh should not count on this, as other competitors are also able to poseserious challenges to its current market position even if Chinese exports are indeedfurther restricted after 2004. The most reliable way of maintaining Bangladesh’scurrent market share and hence reducing its vulnerability to the quota phase-out isto remove the various supply constraints identified in this study and elsewhere.Over the past decades, progress in addressing structural impediments to trade
expansion has been slow. Bangladesh’s poor infrastructure – e.g., its unreliablepower supply, expensive telecommunications networks, congested roads, and ineffi-cient seaports – have imposed substantial costs on its export industry. Together withpoor governance, as reflected in widespread corruption, these have contributed to apoor investment climate which hinders foreign and domestic investment alike.Many of the structural weaknesses of the RMG sector are at least in part policy
induced. After playing a critical role in establishing an exported-oriented RMGindustry, FDI has not been allowed outside of the EPZ sector or to produce quota-restricted products. An important channel of competition, marketing and technologytransfer has been foregone to the detriment of the long-term competitiveness of thedomestic industry. The requirement of back-to-back LCs for input importssignificantly increases the lead times and costs for exporters, as does the reservationof export cargo for domestic vessels.
694 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Given these and other structural weaknesses, Bangladesh faces a serious challengein maintaining its competitiveness in a post-ATC era. This challenge should serve asa wake-up call for policymakers. With its vast labour resources, Bangladesh hasgreat potential to expand its exports of RMGs and other labour-intensive exports ifits key structural weaknesses can be overcome. Bangladesh needs to makedetermined efforts to raise productivity through accelerated structural reforms.
Acknowledgements
We wish to thank, without implication, Nur Calika, Hans Peter Lankes, Olin Liu,Jeremy Mark, Will Martin, Carol Miller, Lynge Nielsen, Zaidi Sattar, DeanSpinanger, and Marijn Verhoeven for their helpful comments, and Ross Arnold,Judith Dean, Vlad Manole, Abul Quasem, Dustin Smith and two anonymousreferees for their valuable assistance with data collection and clarification. The viewsexpressed in this paper are those of the authors and do not necessarily representthose of the IMF or IMF policy.
Notes
1. For a history of trade restrictions on world trade in textiles and clothing, see Keesing and Wolf (1980)
and Bagchi (2001).
2. See Dean (2002) for a succinct summary of how MFA quotas have affected trade patterns across
exporting countries.
3. See the IMF, Market Access for Developing Country Exports – Selected Issues, SM/02/280, http://
www.imf.org/external/np/pdr/ma/2002/eng/092602.pdf.
4. In Bangladeshi usage, the term ‘RMG’ covers items that are not necessarily garments (clothes), such as
towels, napkins, etc. In this paper, the terms ‘T&C’ and ‘RMG’ are used interchangeably, unless the
context demands otherwise.
5. For the history of Bangladesh’s RMG exports, see Spinanger (1987) and Rhee (1990).
6. For a fuller description of the Bangladeshi garment industry, see Bhattacharya (2003), Bhattacharya
and Rahman (2000), and Islam (2001).
7. See Baughman and others (2001) and Centre for Policy Dialogue (2003) for a discussion of the
potential social impact of the quota removal.
8. AGOA provides exclusive quota- and duty-free market access for exports from eligible countries in
Sub-Saharan Africa.
9. Norway’s quotas had been removed for all exporting countries by 2002, while tariffs on exports from
LDCs were eliminated in 2002.
10. If both a quota and tariff are imposed on a product, only one of them can be binding. If the quota is
binding, the tariff simply allows the importing country to recoup some of the quota rent that would
otherwise accrue to the exporting country and the entire preference margin for Bangladesh is the tariff
equivalent of the quota. If the quota is not binding, the entire preference margin for Bangladesh is the
level of the tariff.
11. Average tariffs on textile and clothing in the industrial countries remain high and escalating. They
range from about 8–14 for yarns to 11–22 per cent for clothing. The maximum for clothing goes up to
30 per cent.
12. Under the EU’s so-called ‘SAARC cumulation’ rules, Bangladeshi products made with inputs from
the South Asian Association for Regional Cooperation (SAARC) region (including Bangladesh,
Bhutan, India, the Maldives, Nepal, Pakistan and Sri Lanka) are eligible for preferential treatment
under the EBA, if they meet the minimum value-added requirement.
13. The closure of land ports for textile imports from India in March 2002 has increased difficulties for the
knitwear industry. This policy is supposedly intended to limit smuggling, which the local textile
industry claims to be a major impediment to its efficiency. The true motivation for the policy is
unclear.
The End of Textiles Quotas 695
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
14. See BNA (2004).
15. Cambodia also seems to be extremely dependent on the restricted markets, but reliable data are not
available.
16. For an explanation of how quota removal would affect unrestricted markets, see Martin and
Suphachalasai (1990) and Yang and others (1997).
17. The Finger-Kreinin similarity index is defined as IAB ¼Pn
i¼1 MinðSAi ;S
Bi Þ � 100, where SA
i is the share
of product i in country A’s exports to an export market, SBi is the share of product i in country B’s
exports to the same market, and n is the number of products. It should be noted that the index is
sensitive to the level of product disaggregation. The more disaggregated the products are for the same
level of aggregate exports, the lower the index is. In this study, the index is computed at the quota
level.
18. Other factors affecting quota prices are demand and supply conditions. The growth rate and size of a
quota by themselves do not indicate its restrictiveness unless they are considered together with demand
and supply conditions.
19. The first MFA mandated that quota growth should be no less than 6 percent per year. However,
subsequent extensions of the MFA allowed departure from this growth rate. In general, more
competitive and larger exporters (Hong Kong, Korea, Taiwan, and China) received much lower
growth rates than 6 per cent, while some small, new exporters were given higher growth rates under a
provision that called for higher growth rates for small exporters. The actual growth rates of quotas
may reflect many other factors other than the provision (e.g., overall bilateral trade ties and politics).
20. Under the MFA and ATC, exporting countries are allowed some flexibility in using some of their base
quotas in terms of the so-called carry-forward, carry-over, and swings (shifting a specified proportion
of a quota category to another within the same year). Actual annual growth rates of quotas therefore
may differ from base rates.
21. The demand elasticity depends importantly on the elasticity of substitution between exports from
Bangladesh and its competitors.
22. Note that although these numbers are also computed at the quota level they are not comparable with
those for the US markets as product classifications differ. All results are based on data provided by the
European Commission.
23. See Appendix Table A.8 in World Bank (2004). These ETEs seem to be extraordinarily high, ranging
from 82 per cent to several hundred percent.
24. Data on Indian ETEs at the product level in the EU market are not available.
25. The similarity index failed to capture the potential export competition because binding quotas
on one country raise exports from its competitors while reducing that country’s exports. The result-
ing low export similarity resembles the endogeneity problem of calculating trade-weighted average
tariff.
26. Infrastructure was even important under the quota regime of the MFA. See Andriamananjara and
others (2004).
27. Like in many other countries, quota allocation in Bangladesh was based on past performance. The
system was backward-looking and failed to encourage competition. It favored large and well-
established firms at the expense of newer, potentially more innovative firms.
28. Permission is usually given only for manufacturers that have at least one backward linkage. For a
more detailed discussion of the FDI regime in Bangladesh, see Sattar (2000).
29. A buying house is an intermediary without its own retail operations.
30. Firms may be allowed to import up to four months’ needs if the proposed export complies with norms
under the Import Policy Order, notably minimum value addition of 25 per cent. This generally favors
large firms.
31. For a detailed discussion of Bangladesh’s quota allocation system, see Krishna and Tan (1998).
32. When Sweden joined the EU in 1995 and quotas were reapplied, there was no switch back to
Bangladesh, with other EU rim countries benefiting.
33. Estimates based on the impact of quota removal to date, performance in 10 key garment categories
and discussions with industry leaders.
34. Full documentation of the GTAP model and its companying database can be found in Hertel (1997)
and Dimaranan and McDougall (2002). The GTAP model is solved using the software GEMPACK
(Harrison and Pearson 1996).
35. It is not clear how a fall of exports of 6 per cent can translate to a output decline of 20 per cent.
696 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
36. As noted in Table 9, existing studies tend to use a benchmark database for 1997. The projected 2007
benchmark enables us to capture some of the structural changes other than those arising from the
abolition of MFA quotas. We have also used this exercise to reduce the quota prices for Bangladesh’s
exports to the EU to zero, to reflect the effect of the EBA. The high quota price for Bangladesh’s
textile exports to the U.S. market in the GTAP database was also eliminated in line with our estimates.
Leaving these wedges in the database, as most existing studies do, would distort simulation results,
especially welfare results.
37. It is assumed in the model that quota rents accrue to the exporting country. There is some evidence,
however, that quota rents may be shared between the exporting and importing countries (Krishna and
others 1994).
38. At the margin, this is the unlimited labor supply model. Given the high unemployment rate and the
vast pool of the underemployed in Bangladesh and other developing countries, this assumption is not
unreasonable. Most existing studies tend to assume fixed labor supply, which may not accurately
reflect the fact that layoffs are common, except in the public service sector. The numeraire of the model
is the average global factor price.
39. The results reported are based on the average (8 per cent) of the estimates of quota premiums in
Bangladesh.
40. In the model products are differentiated by country of origin (the Armington assumption). There are
two sets of elasticities of substitution: those between domestic products and imports and those
between imports by country of origin (see Appendix Table A2 for the elasticity values).
41. Even within textile and clothing exports, the product range is limited: eighty percent of exports of
knitwear products are men’s and boys’ shirts, and t-shirts, while 75 per cent of woven products are
men’s and boys’ shirts and trousers and women’s and girl’s trousers. While these products are not
separately modeled, this concentration could bring additional vulnerability to external competition.
42. We could have used Gempack’s systematic sensitivity analysis, but it would not have given us the
direction of change in the results when elasticities varied.
43. The long-term adverse effect of reduced investment on the balance of payments and growth is not
captured here as investment does not affect the capital stock given the static nature of the model.
44. The US has already reimposed quotas on three categories of textile imports from China in December
2003 and is reportedly trying to negotiate a more broadly based new quota system to forestall the
expected rapid expansion of Chinese exports after 2004.
References
Ahmed, S. and Sattar, Z. (2003) Trade liberalization, growth and poverty reduction: the case of
Bangladesh. Paper presented at the World Bank ABCDE Conference, Bangalore, July.
Anderson, K., Dimaranan, B., Hertel, T. and Martin, W. (1996) Asia-Pacific food markets and trade in
2005: a global, economy-wide perspective. Seminar paper 96–05, Centre for International Economic
Studies, University of Adelaide, South Australia.
Andriamananjara, D.C.S., Dean, J. and Spinanger, D. (2004) Trading apparel: developing countries in
2005. Paper presented at the 7th Conference on Global Economic Analysis, Washington, DC, June.
Bagchi, S. (2001) International Trade Policy in Textiles – Fifty Years of Protectionism (Geneva:
International Textile and Clothing Bureau).
Bangladesh Enterprise Institute and World Bank (2003) Improving the Investment Climate in Bangladesh
(Washington, DC: World Bank).
Baughman, L., Francois, J.F. and Spinanger, D. (2001) Estimated effects on United States and Bangladesh
of liberalizing U.S. barriers to apparel imports. Report, Trade Partnership Worldwide, Washington,
DC, November.
Bhattacharya, D. (2003) Final countdown of the MFA: fallout for the LDCs. Centre for Policy Dialogue,
Dhaka, January.
Bhattacharya, D. and Rahman, M. (2000) Experience with implementation of WTO-ATC and
implications for Bangladesh. Centre for Policy Dialogue, Dhaka, September.
BNA (2004) International Trade, 21(12), 18 March, p. 496.
Center for Policy Dialogue (2003) Phasing Out of the Apparel Quota: Addressing Livelihood Concerns
(Dhaka: Center for Policy Dialogue and the University Press).
Cookson, F. (2003a) A short note on export development. Dhaka, March.
The End of Textiles Quotas 697
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Cookson, F. (2003b) Facing 2005 – outlook for the garment sector. Dhaka, March.
Dean, J. (2002) Removing textile and apparel trade barriers: the impact on developing country exporters.
Mimeo, US International Trade Commission, Washington, DC.
Dimaranan, B.V. and McDougall, R.A. (2002) Global Trade, Assistance, and Production: The GTAP 5
Data Base, Center for Global Trade Analysis (West Lafayette, IN: Purdue University).
Dowlah, C.A.F. (1999) The future of the readymade clothing industry of Bangladesh in the post-Uruguay
round world. The World Economy, 22(7), pp. 933–953.
Gereffi, G. (1999) International trade and industrial upgrading in the apparel commodity chain. Journal of
International Economics, 48(1), pp. 37–70.
Gherzi Textile and Project Promotion and Management Associates (2002) Strategic development and
marketing plans for RMG and related industries. Report submitted to the Ministry of Commerce,
Government of Bangladesh, December.
Harrison, W.J. and Pearson, K.R. (1996) Computing solutions for large general equilibrium models using
GEMPACK. Computational Economics, 9, pp. 83–127.
Hertel, T.W. (ed.) (1997) Global Trade Analysis Using the GTAPModel (New York: Cambridge University
Press).
Herok, C.A. and van Tongeren, F. (2002) Bangladesh: textiles versus clothing. Paper presented at ETSG
Conference, Keil, 13–15 September.
Hummel, D. (2001) Time as a trade barrier. Mimeo, Purdue University, July.
Islam, S. (2001) The Textile and Clothing Industry in Bangladesh in a Changing World (Dhaka: Centre for
Policy Dialogue and the University Press).
Keesing, D.B. and Wolf, M. (1980) Textile Quotas Against Developing Countries. Thames Essay No. 23
(London: Trade Policy Research Centre).
Krishna, K., Erzan, R. and Tan, L. (1994) Rent sharing in the multi-fibre arrangement: theory and
evidence from U.S. apparel imports from Hong Kong. Review of International Economics, 2(1), pp.
62–73.
Krishna, K., Erzan, R. and Tan, L. (1998) Rags and Riches: Implementing Apparel Quotas under the Multi-
Fibre Arrangement (Ann Arbor, MI: University of Michigan Press).
Lips, M., Tabeau, A. and van Tongeren, F. (2003) Multilateral and regional trade agreements: options for
Bangladesh. Paper presented at the 6th Conference on Global Economic Analysis, The Hague, 12–14
June.
Martin, W. and Suphachalasai, S. (1990) Effects of the multi-fibre arrangement on developing country
exporters: a simple theoretical framework, in: Carl B. Hamilton (ed.) Textiles Trade and the
Developing Countries: Eliminating the Multi-Fibre Arrangement in the 1990s (Washington, DC: World
Bank), pp. 46–62.
Rhee, Y.W. (1990) The catalyst model of development: lessons from Bangladesh’s success with garment
exports. World Development, 18(2), pp. 333–346.
Sattar, Z. (2000) Foreign direct investment in Bangladesh: issues of long-term sustainability. Policy note,
World Bank Resident Mission, Dhaka.
Spinanger, D. (1987) Will the multi-fibre arrangement keep Bangladesh humble? The World Economy,
10(1), pp. 75–84.
Spinanger, D. and Verma, S. (2003) The coming death of the ATC and China’s WTO accession: will push
come to shove for Indian T&C Exports? Mimeo, Institute for World Economics and OXFAM, Kiel
and New Delhi, June.
Spinanger, D. and Wogart, J.P. (2001) Will the Emperor wear clothes from Bangladesh in 2005? Paper
presented at the IEA Meeting in New Orleans, 5–7 January.
United States International Trade Commission (USITC) (2004) Textiles and Apparel: Assessment of the
Competitiveness of Certain Foreign Suppliers to the U.S. Market. USITC Publication No. 3671,
January (Washington, DC: USITC).
World Bank (2004) Implications for Pakistan of abolishing textile and clothing export quotas. Mimeo,
17 April.
World Trade Organization (WTO) (2000) Trade Policy Review: Bangladesh, Report by the Secretariat
(Geneva: WTO).
Yang, Y., Martin, M. and Yanagishima, K. (1997) Evaluating the benefits of abolishing the MFA in the
Uruguay Round package, in: Thomas Hertel (ed.), Global Trade Analysis Using the GTAP Model
(Cambridge: Cambridge University Press), pp. 253–279.
698 Y. Yang & M. Mlachila
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
Appendix. Data Aggregation and Armington Elasticities in GTAP Model
Table A1. Country/region and industry aggregations used in the model
Country/region Industry
Bangladesh Agriculture and foodUnited States MiningEuropean Union TextilesOther advanced countries ClothingAsian newly industrialized Other manufacturingASEAN ServicesChinaSouth AsiaMiddle East and North AfricaLatin AmericaSub-Saharan AfricaRest of the world
Table A2. Central scenario elasticities of substitution in demand for goods
Commodity
Elasticity of substitutionbetween domesticgoods and imports
Elasticity of substitutionbetween imports bycountry of origin
Agriculture and food 2.4 4.7Mining 2.8 5.6Textiles 2.2 4.4Clothing 4.4 8.8Other manufactures 2.9 6.0Services 1.9 3.9
Source: Based GTAP database Version 5.
The End of Textiles Quotas 699
Dow
nloa
ded
by [
Uni
vers
idad
de
Sevi
lla]
at 0
1:31
13
Nov
embe
r 20
14
