Knowledge Hub · Nº 1 · Apparel Manufacturing Clusters
A Few Dozen Places Dress the World
Global apparel manufacturing is concentrated in a few dozen places. Their factories, suppliers and institutions explain which can survive rising tariffs — and where Guatemala can compete.
The single most decisive force reshaping these clusters is the United States' reciprocal tariff regime: the average duty on American apparel imports reached 35.1 percent in December 2025, a decades-high, up from 14.7 percent the January before. That one number compresses the preference margins that built Dhaka, Phnom Penh, Hawassa and Maseru — and poses a question for Guatemala: which clusters endure, and why?
§ I · Compare Clusters
Compare twenty major apparel clusters by exports, employment and production capabilities. Most export and employment figures cover whole countries and are marked “nat'l”; cluster-level estimates are approximate.
Filter clusters
Origin
How did it come to exist?
Organic districts (Prato, Tiruppur, Ludhiana, Gamarra) grew from trade and kin; planned ones were legislated — export-processing and special economic zones, Qualifying Industrial Zones, bonded-warehouse and maquila regimes, and purpose-built industrial parks.
Production model
What does a factory sell?
The upgrading ladder: cut-make-trim on the buyer's fabric → full-package (OEM) sourcing its own → design for the buyer (ODM) → the factory's own brand (OBM). Most of the world's apparel still sits on the first two rungs.
Vertical integration
Does the fabric come from next door?
The decisive variable for rules of origin. Yarn-forward (CAFTA-DR) and double-transformation (EU GSP+) rules reward clusters with a textile base next door and punish those sewing imported cloth.
Market positioning
What is it known for?
Ownership structure
Whose capital?
Foreign direct investment built Vietnam, Cambodia and Hawassa; domestic dynasties built Dhaka, Tiruppur and Colombo's groups; immigrant entrepreneurs remade Prato and Los Angeles.
Lifecycle stage
Where on the curve?
A cluster can grow, mature, decline or renew. Stages can overlap: Istanbul is mature and pressured; Prato mature and reinventing itself.
Scale
How big?
| # | Cluster | Country | Export value (year) | Employment | Firms | Lifecycle / trend |
|---|---|---|---|---|---|---|
| 1 | Guangdong complexHumen · Xintang · Panyu · Shaxi | China | nat'l US$165.24bn apparel · 2024 | Humen 200,000+ · Xintang ~220,000 | Humen ~3,000 · Xintang 2,600+ | — Mature / stable |
| 2 | Zhejiang complexKeqiao · Datang · Zhili · Pinghu | China | Keqiao market RMB 441bn · 2025 | Datang ~100,000 | Keqiao ~8,000 mfrs · 22,000 in market | — Mature / stable |
| 3 | Greater DhakaGazipur · Savar/Ashulia · Narayanganj | Bangladesh | nat'l US$38.48bn RMG · 2024 | ~4m nat'l | ~3,500–4,000 active | ▲ Growing · at risk (LDC 2026) |
| 4 | ChattogramCEPZ · Karnaphuli EPZ | Bangladesh | part of nat'l total | — | — | ▲ Growing |
| 5 | Vietnam Southern hubHCMC · Binh Duong · Dong Nai · Long An | Vietnam | part of nat'l ~US$46bn · 2025 | ~3m nat'l | 3,500 FDI projects nat'l | ▲ Growing (China+1) |
| 6 | Vietnam Northern hubHanoi · Hung Yen · Nam Dinh · Bac Giang | Vietnam | — | — | — | ▲ Growing |
| 7 | TiruppurTamil Nadu | India | ₹44,747 cr (~US$5.39bn) · FY2024-25 | ~600,000 direct | 1,360 TEA member units | ▲ Growing |
| 8 | Istanbul complexIstanbul · Bursa · İzmir · Denizli | Turkey | nat'l US$17.9bn RTW · 2024 | 738k → <564k (2022 → Jan 2025) | — | ▼ Declining / pressured |
| 9 | Western ProvinceKatunayake · Biyagama · Koggala | Sri Lanka | nat'l US$4.76bn · 2024 | 350,000+ | ~300 · MAS / Brandix / Hirdaramani | — Stable / resilient |
| 10 | Punjab / Sindh beltFaisalabad · Karachi · Lahore · Sialkot | Pakistan | nat'l US$17.88bn T&A · FY2024-25 | — | 442 spinning + garment units | ▲ Growing |
| 11 | Phnom Penh / KandalPhnom Penh · Kandal · Kampong Speu | Cambodia | nat'l GFT >US$9.33bn · Jan–Oct 2024 | 750,000+ | 580+ GMAC facilities | ▲ Growing · EBA-dependent |
| 12 | West / Central JavaBandung · Jakarta · Tangerang | Indonesia | nat'l (large) | ~ millions | — | — Stable |
| 13 | Ave-Cávado (Norte)Porto · Braga · Guimarães · Famalicão | Portugal | €6.1bn · 2024 (~80% of nat'l) | ~130,000 | ~12,000 (99.7% SMEs) | ▲ Renewing (nearshoring) |
| 14 | Sula ValleySan Pedro Sula · Choloma · Villanueva | Honduras | CAFTA-DR region >US$11bn · 2024 | ~146,000 direct (AHM) | 19 textile + maquilas | — Stable (nearshoring) |
| 15 | ◆ Guatemala City metroGuatemala City · Sacatepéquez | Guatemala | part of CAFTA-DR >US$11bn · 2024 | ~ tens of thousands | Vestex members | ▲ Growing (Nov 2025 CAFTA deal) |
| 16 | PratoTuscany | Italy | ~€4.2bn turnover · 2024 | ~32,000 direct | ~5,000–6,800 · mostly Chinese-run | — Mature / contested |
| 17 | Los Angeles Fashion DistrictLos Angeles | USA | — | ~ tens of thousands | thousands (small) | ▼ Declining / pressured |
| 18 | Medellín / AntioquiaMedellín · Bogotá | Colombia | nat'l · Colombiatex / Colombiamoda | — | — | — Stable |
| 19 | Hawassa Industrial ParkHawassa · Bole Lemi · Adama | Ethiopia | park ~US$114m/yr (peak) | peak ~35,000 → ~3,000 | ~20 firms (post-exodus) | ▼ Declining (AGOA loss) |
| 20 | MaseruMaseru | Lesotho | nat'l ~US$237m to US (pre-shock) | ~40,000 | — | ▼ Declining · AGOA reprieve to 2028 |
SOURCE: WTO VIA FASH455 · EPB/BGMEA · VITAS · TEA · İHKİB · JAAF · PBS · GMAC · ATP · AHM · VESTEX · CNTAC · XINHUA/CHINA DAILY (CHINA FIGURES SELF-REPORTED) · RANKED BY STRATEGIC SIGNIFICANCE + EXPORT WEIGHT
More manufacturing centres
Xintang, Guangzhou CN · DENIM
2,600+ denim enterprises; ~300 million pairs a year, ~60% of national output; ~220,000 workers — figures of 2011–18 vintage, repeated since.
Datang, Zhuji CN · SOCKS
~25 billion pairs a year — over 70% of China's output and a third of the world's; 6,000+ sock enterprises; output value ~RMB 75bn.
Zhili, Huzhou CN · CHILDREN'S WEAR
14,000+ enterprises and 8,000+ e-commerce firms; ~1.5 billion garments a year; 2022 sales above RMB 70bn, about two-thirds of China's kids-wear market.
Jinjiang / Quanzhou CN · SPORTSWEAR · OBM
Footwear + apparel output above RMB 300bn (2022); ~40% of China's sports shoes, ~20% of the world's; home of Anta, Xtep and 361°.
Keqiao, Shaoxing CN · FABRIC MARKET
The world's largest fabric market: RMB 400.99bn (2024) → RMB 441.39bn (2025); about a quarter of world fabric traded here, to 206 countries.
Mauritius MU · ODM/OBM · DECLINING
Textile + apparel exports ~US$726m; Europe's fourth T-shirt supplier; 2024 clothing exports −14% as costs push production to Madagascar.
Antananarivo EPZ MG · EPZ
The largest EPZ employer, 90,000+ workers; 20–30% of export revenue; AGOA access lost, regained, and hit again by 2025 US tariffs.
Al-Hassan · Ad-Dulayl · Sahab JO · QIZ
US-FTA and QIZ-driven; large migrant-labour share; ILO Better Work Jordan.
Greater Cairo · Alexandria · Port Said EG · QIZ
QIZ access to the United States; growing.
SONAPI · Caracol · CODEVI HT · CMT · DECLINING
Apparel ~90% of exports; jobs fell from 60,000+ (2021) to ~22,000 (2024); US imports US$549m in 2024, 66% under HOPE/HELP.
Torreón · Tehuacán · Aguascalientes MX · DENIM · MAQUILA
NAFTA-era boom, MFA-phase-out bust; partial nearshoring recovery under USMCA yarn-forward.
Lima — Gamarra & Pima exporters PE · PREMIUM KNIT · ORGANIC
Premium organic and Pima-cotton knitwear niche; Gamarra is the organic district, the exporters the OBM tier.
Santa Catarina · São Paulo BR · DOMESTIC MARKET
Blumenau/Jaraguá and Brás/Bom Retiro; largely oriented to the home market.
Tangier · Casablanca · Monastir · Sfax MA / TN · NEARSHORE
Fast-turnaround nearshoring to the EU for Inditex and Spanish and French buyers.
◆ Guatemala, benchmarked
The Only Peer With a Zero in the Tariff Column
Read across the row and the Guatemalan cluster's case is arithmetic. The 13 November 2025 US–Guatemala framework restored duty-free CAFTA-DR access and removed the reciprocal tariff; every Asian competitor now pays 19 to 50 percent at the American border. What Guatemala lacks is in the third column: its fabric still largely arrives by ship, and yarn-forward rules reward the neighbour that spins its own.
Full-package capability and the Vestex–AGEXPORT ecosystem place the cluster on the second rung of the upgrading ladder, one rung ahead of Honduras's basics — and one rung behind Sri Lanka's design houses.
GT Benchmark · Nearshore and Asian peers
STATE AS OF SOURCE DATES · 2025–26
| Cluster | Model | Preference · RoO | US tariff · Aug 2025 → | Trend |
|---|---|---|---|---|
| ◆ Guatemala City metro | Full-package | CAFTA-DR · yarn-forward | duty-free · 13 Nov 2025 framework | ▲ growing |
| Sula Valley | Full-package | CAFTA-DR · yarn-forward | framework pending | — stable |
| Vietnam Southern hub | Full-package | CPTPP · EVFTA | 20% (from 46% announced) | ▲ growing |
| Greater Dhaka | CMT → full-package | EU EBA · to Nov 2029 | 20% (from 37%) | ▲ at risk |
| Tiruppur | Full-package | UK FTA · 27 Jul 2025 | 50% (from 26%) | ▲ growing |
| Phnom Penh / Kandal | CMT → full-package | EU EBA · partial | 19% | ▲ growing |
| Western Province | ODM | — | 44% announced · negotiated down | — stable |
| Istanbul complex | ODM / OBM | EU customs union | — | ▼ pressured |
| Ave-Cávado (Norte) | ODM | EU single market | — | ▲ renewing |
AVERAGE US APPAREL-IMPORT TARIFF 35.1% (DEC 2025) VS 14.7% (JAN 2025). RATES ARE MOVING TARGETS; ROO = RULES OF ORIGIN. HONDURAS, DR AND COSTA RICA FRAMEWORKS PENDING AT COMPILATION.
§ II · How Clusters Grow, Decline and Renew
Trade preferences can launch a cluster; stronger suppliers, skills and institutions help it endure. Growth, maturity and renewal can overlap, while rising costs or lost market access can force a decline.
01
Emergence
A trade preference — AGOA, EBA, MFA quota-hopping, a QIZ — or a diaspora seed. Entry at CMT assembly on imported fabric and low wages; FDI or domestic entrepreneurs supply the first firms.
Example: Uzbekistan
02
Growth
Labour pooling and supplier entry deepen; associations form; export volumes rise; some backward integration into fabric begins.
Greater Dhaka · Chattogram · Vietnam Southern hub · Vietnam Northern hub · Tiruppur · Punjab / Sindh belt · Phnom Penh / Kandal · ◆ Guatemala City metro
03
Maturity & upgrading
Full-package, then ODM, occasionally OBM. Process upgrading (automation), product upgrading (unit value), functional upgrading (design, logistics). Sri Lanka, Turkey and Mauritius are the textbook cases.
Guangdong complex · Zhejiang complex · Istanbul complex · Western Province · West / Central Java · Sula Valley · Prato · Medellín / Antioquia
04
Lock-in & decline
Preferences erode (Torreón after the MFA), wages rise without productivity (Mauritius, Turkey), or a shock lands — AGOA suspended in Ethiopia, the coup in Myanmar, state collapse in Haiti. Firms that never upgraded are locked in.
Istanbul complex · Los Angeles Fashion District · Hawassa Industrial Park · Maseru
05
Renewal
Repositioning on speed, sustainability and proximity. Portugal's Norte abandoned price competition for premium nearshoring; Prato's diaspora reinvented the district as Europe's pronto-moda hub.
Ave-Cávado (Norte) · Prato
Success factors · what the survivors share
- 01A trade preference plus achievable rules of origin — CAFTA-DR yarn-forward, EBA, AGOA
- 02A local textile base — China, Turkey, Sri Lanka's Eravur push, Central America's Elcatex
- 03Strong associations and collective action — TEA's zero-liquid discharge, BGMEA/RSC, JAAF, İHKİB, GMAC, AHM, Vestex
- 04Industrial parks with shared infrastructure, above all effluent treatment
- 05Skills and technology institutions — CITEVE, the design schools
- 06Lead-firm anchoring — Inditex in Galicia, PVH at Hawassa, Anta in Jinjiang
- 07Proximity and speed to market — Turkey, Portugal, Central America, Morocco
- 08Entrepreneurial and diaspora networks — Prato's Chinese firms, Tiruppur's Gounders, Los Angeles
- 09Compliance and sustainability as differentiation — Sri Lanka, Bangladesh's green factories, Portugal
Failure modes · how clusters die
- †Dependence on a single market or a single preference
- †Failure to upgrade beyond CMT
- †Wage inflation without productivity gains
- †Environmental non-compliance
- †Political instability
- †Buyer consolidation and flight
The model transfers where a country combines a stable preference with workable rules of origin, either FDI or domestic entrepreneurs, associations able to provide collective goods, and infrastructure. Absent any one, clusters stall.
§ III · Lessons From Twelve Clusters
Tiruppur turned an environmental shutdown into a competitive advantage. Hawassa lost its anchor buyer when trade preferences disappeared. These twelve cases show what sustains a cluster — and what can undo it.
India · Tiruppur
Tiruppur: the district that turned a shutdown into a sales pitch
A Marshallian knitwear district built by the Gounder farming community; ~600,000 workers and 1,360 exporting units in a compact geography of knitting, dyeing, printing and stitching. A 2011 court-ordered dyeing shutdown forced zero-liquid-discharge on the whole cluster — now a marketing asset. Boosted by orders shifting out of Bangladesh in 2024.
- EXPORTS
- ₹44,747 cr (~US$5.39bn) FY24-25 · ▲ +16.49% · 68.65% of India's knitwear
- RISK
- US 50% tariff on India, Aug 2025 — harshest in Asia
- LESSON
- Collective environmental action wins; preference gaps vs neighbours decide
Bangladesh · Dhaka
Greater Dhaka: safety rebuilt confidence, one preference carries the risk
Seeded by Daewoo–Desh in 1978 and MFA quota-hopping; the world's second exporter, still fabric-import-dependent for wovens. After Rana Plaza (2013) the Accord/Alliance, now the RMG Sustainability Council, inspected 1,600+ plants; Bangladesh leads the world in LEED-certified factories.
- EXPORTS
- US$38.48bn 2024 · US$39.35bn FY24-25 · EU ~50%
- RISK
- LDC graduation Nov 2026 → EBA ends Nov 2029; up to US$8bn/yr at stake
- LESSON
- Compliance preserved buyers; single-preference dependence is the vulnerability
Vietnam
Vietnam: the model China+1, with a fabric gap
Korean, Taiwanese and Chinese FDI plus CPTPP and the EU–Vietnam FTA, on a foundation of political stability. Domestic value-added has reached ~52%; yarn and fabric are still partly imported from China. Target: US$64.5bn by 2030.
- EXPORTS
- ~US$44bn 2024 · crossed US$46bn 2025 · US ~US$18.6bn
- RISK
- US tariff 46% announced → 20% negotiated; CPTPP yarn-forward gaps
- LESSON
- FTAs + FDI + stability; the fabric supply gap is the upgrading frontier
Sri Lanka
Sri Lanka: garments without guilt, at a premium
Anchored by MAS, Brandix and Hirdaramani — 350,000+ workers, ~6% of GDP — the cluster pivoted to quality and ethics after the MFA and now co-develops lingerie and activewear with Victoria's Secret, Nike and lululemon. The Eravur Textile Zone is the bid for local fabric.
- EXPORTS
- US$4.76bn 2024 · ▲ +5% · ~US$5bn 2025 · still 10.3% below 2019
- RISK
- US tariff 44% announced; energy costs; Colombo port bottlenecks
- LESSON
- Compliance + functional upgrading = durable premium positioning
Portugal
Portugal Norte: a declining district that stopped competing on price
~12,000 firms (99.7% SMEs), ~130,000 workers, a near-complete chain from spinning to finishing, CITEVE for technology and ATP for advocacy. Nearshoring after the pandemic and the Red Sea, EU sustainability rules and low-MOQ direct-to-consumer demand did the rest.
- EXPORTS
- €6.1bn 2024 · ~80% of national output · 3rd EU exporter
- RISK
- Wage inflation; ageing workforce; vocational shortfall
- LESSON
- Renewal through speed, quality, sustainability and proximity
Turkey
Istanbul: the vertical chain the lira could not protect
Cotton to garment in one country, an hour's flight from its buyers — 60% of exports go to the EU. Then currency volatility and a cost–exchange-rate imbalance eroded competitiveness; imports rose to ~US$4bn.
- EXPORTS
- US$21.2bn peak 2022 → US$17.9bn 2024 · ▼
- JOBS
- 738,000 (end-2022) → under 564,000 (Jan 2025)
- LESSON
- Macro instability can override every cluster strength
Honduras · Sula Valley
Sula Valley: the yarn-forward template
CAFTA-DR duty-free access and proximity to the United States, under Decree 29-89. ~146,000 direct textile jobs; Elcatex (knit fabric, full-package, 6,000 staff), Gildan, Fruit of the Loom and Grupo Karim's integrate the chain regionally; AHM leads on solar and low-water dyeing.
- EXPORTS
- CAFTA-DR region >US$11bn 2024 · ▼ −8% in 2025 on reciprocal tariffs
- RISK
- Low-skill workforce; little upgrading beyond basics; US dependence
- LESSON
- Yarn-forward rules + regional fabric = a genuine US–Central America chain
◆ Guatemala
Guatemala: association-led branding on the second rung
Guatemala City metro and Sacatepéquez; knitwear and sportswear sold full-package under "Made in Guatemala", with a skilled workforce and the Vestex–AGEXPORT ecosystem behind it. The November 2025 US framework restored duty-free CAFTA-DR access.
- TARIFF
- Duty-free · framework of 13 Nov 2025
- RISK
- Fabric largely imported; US-market dependence
- LESSON
- Full-package capability + a branding association differentiate it from lower-skill neighbours
Italy · Prato
Prato: an immigrant reinvention with a social bill
A post-war recycled-wool district remade from the 1990s by Chinese entrepreneurs: Italian firms upstream in quality and recycled textiles, ~5,000 mostly small Chinese-run pronto-moda subcontractors downstream. Nearly a quarter of residents are of Chinese heritage.
- TURNOVER
- ~€4.2bn 2024 · ~32,000 direct workers
- RISK
- Labour and safety violations; customs fraud; 2024–25 "hanger wars"
- LESSON
- Diaspora capital regenerates; informality and weak enforcement bill the district
Cambodia
Cambodia: monitored compliance as market access
EU Everything-But-Arms access — garments, footwear and travel goods are ~75% of merchandise exports — and ILO Better Factories Cambodia, the pioneer of the Better Work model. 580+ GMAC facilities; 750,000+ workers, mostly women.
- EXPORTS
- GFT >US$9.33bn Jan–Oct 2024
- RISK
- 2020 partial EBA withdrawal; minimum wage US$80 (2013) → US$204 (2024); US 19%
- LESSON
- Transparent compliance is an asset; wages and conditionality force upgrading
Mauritius
Mauritius: upgrading bought longevity, not immortality
The 1970s EPZ pioneer, with 37 years of EU duty- and quota-free access and Hong Kong capital, climbed to fully-fashioned knitwear and its own brands (CMT, Lovable). A small, high-cost island eventually offshores its own low-value tiers — to Madagascar.
- EXPORTS
- Textiles ~US$726m · clothing ▼ −14% in 2024
- RISK
- Cost inflation; AGOA uncertainty; 2025 US punitive tariff
- LESSON
- Upgrading extends the life cycle; it does not repeal it
Ethiopia · Hawassa
Hawassa: a satellite platform loses its planet
A flagship state-anchored eco-park anchored by PVH and Asian suppliers, peaking at ~35,000 workers. AGOA was suspended in January 2022 over the Tigray war; PVH left; 18 foreign firms exited, ~11,500 jobs and US$45m went with them. Employment fell to ~3,000.
- STATUS
- AGOA reinstatement repeatedly denied; pivot to the EU under way
- RISK
- One preference, one anchor tenant
- LESSON
- A planned park is only as strong as its single preference and its politics
Cautionary cases · what decline looks like
| Case | What happened | Lesson |
|---|---|---|
| Torreón / La Laguna denim (Mexico) | NAFTA-era boom, bust after the MFA phase-out and China's WTO accession; partial USMCA nearshoring recovery | Preference-dependent booms reverse when the preference is competed away |
| Lesotho & Eswatini | April 2025 US tariff (50% on Lesotho) + 30 Sep 2025 AGOA lapse → cancelled orders, closures, national state of disaster; AGOA restored to 2028, buyers not returning | Single-market dependence transmits shocks instantly to workers, mostly women |
| Myanmar (Yangon) | Post-2021 coup; EBA-withdrawal risk; buyer exits | Political risk overrides low-cost advantage |
| Haiti | Apparel ~90% of exports; jobs 60,000+ (2021) → ~22,000 (2024); HOPE/HELP lapsed 1 Oct 2025, restored Feb 2026; Hurricane Melissa | Trade preference cannot offset state collapse |
| Leicester (UK) | Boohoo supply-chain scandal — sub-minimum wages, unsafe conditions exposed 2020 | Domestic clusters can harbour sweatshop conditions |
| US Southeast & Northern England | Mass offshoring 1990s–2000s; remnants in technical, yarn and automated segments | Renewal requires moving up-market or into capital-intensive niches |
| Madagascar | Lost then regained AGOA (2010s); 2025 US punitive tariffs hit again | Volatile preference access deters long-term investment |
§ IV · Outlook 2024–2026 · The Tariff Is the Map
The 2 April 2025 "Liberation Day" schedule, a ninety-day pause and a summer of country deals reset the rates from 7 August 2025. The net effect is perverse for the China+1 strategy: a brand that left China now finds virtually no tariff benefit in Vietnam or Bangladesh, and India — at 50 percent including a 25-point penalty over Russian oil — is the biggest loser among Asian suppliers. Meanwhile de minimis closed for China and Hong Kong on 2 May 2025, and then more broadly.
30 SEP 202503 FEB 202602 SEP 2026
AGOA — expired, reinstated, extended
Lapsed; reinstated retroactively to 31 December 2026; then extended to 31 December 2028 in the continuing appropriations act. Section 122 surcharges still apply; AGOA exports were down ~32% in the year to November 2025. A reprieve, not a resolution — the buyers who left Lesotho and Madagascar have largely not returned.
NOV 2026NOV 2029
Bangladesh graduates from LDC status
A three-year EBA transition to November 2029, then standard GSP (~9.6% MFN on knit and woven) unless GSP+ — double-transformation rules of origin, safeguard thresholds — or an FTA is secured. WTO and IGC estimate losses of up to US$8bn a year, ~14% of export earnings.
01 OCT 2025FEB 2026
Haiti's HOPE/HELP lapses and returns
Lapsed during the US government shutdown; tariffs of 14–30% hit; restored retroactively through 2026 with duty refunds. The sector had already halved.
18 MAR 202619 JUL 2026~JUN 20272028–2029
The EU rewards the traceable and the proximate
Omnibus I narrowed CSRD/CSDDD to very large firms but left ESPR and the Digital Product Passport untouched; the textiles delegated act is expected late 2026–2027, DPP compliance realistically 2028–2029. Destroying unsold apparel is banned for large firms from 19 July 2026; textile EPR is mandatory EU-wide by ~June 2027; France's environmental-cost label starts October 2026. Portugal, Turkey and Sri Lanka are positioned to benefit.
STRUCTURAL
Nearshoring and China+1, slowly automating, increasingly hot
Central America under the restored CAFTA frameworks; Turkey, Morocco and Tunisia for the EU; Portugal and Eastern Europe for premium; Vietnam and India as growth engines; Africa clouded by AGOA. Automation and on-demand manufacturing advance slowly; heat, flooding and Hurricane Melissa make climate a cluster-strategy variable.
US reciprocal tariffs · apparel origins
ANNOUNCED → RESET
AVG US APPAREL TARIFF · DEC 2025
35.1%
JAN 2025 · BEFORE
14.7%
| Origin | 2 Apr 2025 · announced | 7 Aug 2025 · reset |
|---|---|---|
| Vietnam | 46% | 20% |
| Bangladesh | 37% | 20% |
| Cambodia | 32–49% | 19% |
| Indonesia | 32–49% | 19% |
| Pakistan | — | 19% |
| Sri Lanka | 44% | negotiated down |
| India | 26% | 50% † |
| China | 34%+ | — |
| Lesotho | 50% | 50% · AGOA to 2028 |
DE MINIMIS CLOSED FOR CN/HK 2 MAY 2025, THEN BROADENED. RATES REFLECT SOURCE DATES.
§ V · Priorities for Buyers and Policymakers
For sourcing brands and buyers
NOW0–6 MONTHS
Map tariff exposure by SKU and origin under the December 2025 US schedule. Prioritise CAFTA-DR yarn-forward programmes — Guatemala and El Salvador restored in November 2025 — for US-bound knit basics and fleece; avoid over-concentration in India for US programmes. REVISIT IF ANY US RATE MOVES >5 POINTS ON A TOP-3 ORIGIN.
NEAR TERM6–18 MONTHS
For EU-bound volume, weight toward vertically integrated, traceable clusters — Turkey, Portugal, Sri Lanka — ahead of ESPR and the passport; start collecting fibre composition and country of manufacture now. Model Bangladesh past 2029 under a 0–12% EU duty scenario.
STRUCTURAL18–36 MONTHS
Build dual-region redundancy: one Asian full-package hub, one nearshore hub. CHANGE STRATEGY ON: GSP+ CONFIRMATION FOR BANGLADESH, OR A DURABLE MULTI-YEAR AGOA/CAFTA SETTLEMENT.
For cluster policymakers and development agencies
LEVER 01
Invest in shared infrastructure first — effluent treatment and zero-liquid discharge (the Tiruppur lesson), reliable power, bonded logistics. The highest-leverage collective good.
LEVER 02
Fund backward integration into fabric to satisfy yarn-forward and double-transformation rules — the decisive constraint for Bangladesh, Central America and Cambodia.
LEVER 03
Strengthen associations and compliance institutions on the RSC/Better Work model, and skills and design institutes that make functional upgrading possible.
LEVER 04
Diversify preferences and markets before the shock, not after it. Ethiopia, Lesotho and Haiti are the cost of single-market dependence, itemised.
§ VI · The Research Behind the Analysis
Marshall explains why firms gather in places like Tiruppur; Gereffi explains why manufacturers earn so little of the retail price; Menzel and Fornahl explain why clusters grow, decline and renew.
Why firms cluster
MARSHALL · 1890
Agglomeration economies
Co-location pools labour, breeds specialised suppliers and lets knowledge spill over — "the secrets of industry are in the air". Tiruppur, Prato, Keqiao.
PORTER · 1990, 1998
The competitive cluster
Proximate, interconnected firms and institutions; competitiveness from factor conditions, demand, related industries and rivalry — the "diamond".
BECATTINI · 1979–90
The Italian industrial district
The district as a socio-territorial entity — community plus firms; the basis of "Made in Italy".
PIORE & SABEL · 1984
Flexible specialisation
Small-firm networks out-competing mass production through flexibility — the Prato and Carpi model.
MARKUSEN · 1996
Four cluster types
Marshallian (many SMEs), hub-and-spoke (Inditex in Galicia), satellite platform (Hawassa's branch plants), state-anchored.
SCHMITZ · 1995
Collective efficiency
Advantage = passive agglomeration + active joint action through associations. Why TEA's effluent plants matter.
Who governs the chain
GEREFFI · 1994–
Global value chains
Apparel is the archetypal buyer-driven chain; upgrading runs process → product → functional → inter-sectoral.
HUMPHREY & SCHMITZ · 2002
Governance and upgrading
Local upgrading is conditioned by chain governance; the risk is being locked into low-value functions.
GEREFFI, HUMPHREY & STURGEON · 2005
Five governance types
Market, modular, relational, captive, hierarchy — most apparel sits in captive or modular chains.
HENDERSON · COE · DICKEN · HESS
Global production networks
Broader than chains: embeddedness in territory and institutions.
EMPIRICAL GVC LITERATURE
Learning by exporting
Productivity and quality gains from serving demanding global buyers.
COOKE · ETZKOWITZ & LEYDESDORFF
Regional innovation systems
The triple helix — university, industry, government: design schools and testing labs as cluster institutions.
How clusters change, and how states make them
MENZEL & FORNAHL · 2010
The cluster life cycle
Emergence → growth → maturity → decline or renewal, driven by the heterogeneity of knowledge.
MARTIN & SUNLEY · 2006
Path dependence and lock-in
Clusters rigidify; negative lock-in sets in when preferences erode or wages rise. Torreón.
FRENKEN · BOSCHMA
Related variety
Diversification into technologically related activities aids renewal — knitwear into technical textiles, denim into laundries.
RECENT POLICY LITERATURE
Nearshoring and friend-shoring
Proximity plus geopolitical alignment reshaping sourcing — Central America, Turkey, Portugal, North Africa, Eastern Europe.
UNIDO · WORLD BANK · UNCTAD
Planned clusters
State-created agglomerations — industrial parks, bonded zones, QIZs — as policy substitutes for organic emergence.
§ VII · Glossary and Sources
- AGOA
- African Growth and Opportunity Act — US preference for sub-Saharan Africa
- CMT
- Cut-Make-Trim — assembly on the buyer's fabric
- OEM/ODM/OBM
- Original Equipment / Design / Brand Manufacturing — the upgrading ladder
- RMG
- Ready-Made Garment
- EPZ/SEZ/QIZ
- Export Processing / Special Economic / Qualifying Industrial Zone
- EBA
- Everything But Arms — EU duty-free access for LDCs
- GSP/GSP+
- Generalised Scheme of Preferences; the plus requires double transformation
- CAFTA-DR
- Central America–Dominican Republic FTA; yarn-forward rules of origin
- HOPE/HELP
- Haiti trade-preference acts
- ZLD/CETP
- Zero Liquid Discharge / Common Effluent Treatment Plant
- RSC
- RMG Sustainability Council — successor to the Bangladesh Accord
- ESPR · DPP
- Ecodesign for Sustainable Products Regulation · Digital Product Passport
- CSRD/CSDDD
- Corporate Sustainability Reporting / Due Diligence Directive
- EPR · LDC
- Extended Producer Responsibility · Least Developed Country
- GVC
- Global Value Chain
- CNTAC
- China National Textile and Apparel Council
- CPTPP · EVFTA · RCEP
- Trade agreements anchoring Vietnam's access
- HS 61 · 62 · 63
- Knitted · woven · other made-up apparel and textiles
- Associations
- BGMEA/BKMEA · VITAS · TEA/AEPC · JAAF · GMAC · AHM · Vestex/AGEXPORT · İHKİB/İTKİB · ATP · MEXA/EDB
Caveats
† Cluster-level figures are approximations; most authoritative data are national, and sub-national attributions vary between sources and years.
‡ China cluster figures are heavily self-reported (Xinhua, local portals, CNTAC); some are 2011–2018 vintage repeated in later press. Treat as order-of-magnitude.
§ 2025–2026 tariff rates were negotiated repeatedly; rates reflect source dates. Turkey, Portugal and China totals mix "ready-to-wear", "apparel" and "textile + apparel" definitions.
¶ LDC-graduation losses, ESPR/DPP dates and reshoring projections are projections, not realised outcomes.
Principal sources
WTO clothing and textile shares 2024 via Dr Sheng Lu, FASH455 · OTEXA/USITC tariff analysis, FASH455 · EPB/BGMEA via BSS and Textile Focus · VITAS via VnEconomy · TEA 35th AGM and Business Standard · JAAF via Just-Style and HGI · İHKİB 2024 annual report via Kohan · ATP/EURATEX · PBS via Fibre2Fashion · GMAC via Open Development Cambodia · FUNDESA benchmark 2024; Sobel on the CAFTA framework · NBE/AfDB via Capital Ethiopia; IGC · Brownstein, ISS Africa, ODI on AGOA · Daily Star, IGC on LDC graduation · USITC via FASH455 and Haitian Times on HOPE/HELP · China Observers, FashionNetwork on Prato · Baidu Baike, Xinhua, China Daily, Global Times, CNTAC on Chinese clusters · Carbonfact, ESPR Atlas on EU regulation.