The Tariff Cliff

Executive Summary. The EU and UK together absorb 61.6 percent of Bangladesh's total goods exports (51.1 percent and 10.5 percent respectively, EPB FY2024). The Everything But Arms preference that makes this trade duty-free lapses on November 24, 2029, when the three-year grace window expires. After that date, the EU's most-favoured-nation tariff schedule applies: 9.6 percent on knit apparel, 12 percent on woven. The only successor arrangement that fully preserves duty-free access is GSP+, which requires Bangladesh to ratify and credibly implement seven outstanding ILO core conventions. The filing deadline that allows the EU's twelve-to-eighteen-month institutional process to complete before November 2029 is Q2 2027, eighteen months from now. The window is open; the clock is running.

In a glass-walled meeting room at the Bangladesh Mission in Brussels in early May 2026, a junior commercial officer was preparing a one-page brief for the ambassador on what the European Commission's GSP+ ratification track requires. The brief lists twenty-seven international conventions. Eight of them already ratified and implemented. Twelve ratified but implementation patchy. Seven not ratified at all. The seven include several of the International Labour Organization core conventions on freedom of association in export processing zones. The brief notes, in one line at the bottom, that the EU Parliament's Trade Committee is unlikely to approve a GSP+ designation for a country with unratified freedom-of-association conventions and that the Commission has signalled this publicly.

In a knit garment factory off the Mirpur DOHS road on the same morning, the production manager was finishing a costing for a buyer's autumn-winter order that had previously been ten percent of the plant's annual output. The buyer's regional sourcing office had cut the order to four percent and rerouted the remaining six percent to a Vietnamese supplier. The buyer's stated reason was capacity reallocation. The buyer's actual reason was that the order would land in 2030, by which time the Bangladesh shipment would carry a 9.6 percent EU MFN tariff and the Vietnamese shipment would carry zero under the EU-Vietnam FTA.

The Brussels brief and the Mirpur costing are the same problem, with three and a half years between them, and a grace window that is closing while we treat it as time to relax.

What graduation actually changes

On November 24, 2026, Bangladesh is currently scheduled to exit the United Nations' Least Developed Country category. The trigger criteria are the three-year average per-capita gross national income (above the threshold), the Human Assets Index (above the threshold), and the Economic Vulnerability Index (below the threshold). The Committee for Development Policy confirmed our graduation in 2021 and reaffirmed it after pandemic-period reviews. The deferral built into the 2021 decision is the three-year transition that runs until November 2029 for most preferential market-access arrangements. The Tarique Rahman government has subsequently filed a request with the Committee for Development Policy seeking a further three-year deferral of the graduation date itself to November 2029, on which a decision is still pending; this piece treats the current calendar as the binding case for planning purposes.

Two multilateral and bilateral commitments anchor the trade-preferences side of the calendar. The WTO Ministerial Conference in 2024 formally agreed to give graduating LDCs three additional years of preference access. The European Union, the United Kingdom, Canada, and Australia have each publicly committed to extending duty-free access for Bangladesh through November 2029. These commitments are what convert the formal grace into operational predictability.

The technical effects, in plain terms, are these.

The European Union's Everything But Arms scheme grants duty-free quota-free access on all products except arms. As of FY2024 figures published by the Export Promotion Bureau, the EU absorbs 51.1 percent of Bangladesh's total goods exports, and the United Kingdom absorbs another 10.5 percent. Together that is 61.6 percent of our export book. Within the EU, the bulk of our shipments are knit and woven apparel, where EBA preference saves between roughly 9.6 percent (knit) and 12 percent (woven) per dollar of landed value relative to the EU's most-favoured-nation tariff schedule. After November 2029, that preference lapses. Without a successor arrangement, the same shipments land in Rotterdam or Hamburg paying full MFN duty.

The next-best arrangement available to a graduated lower-middle-income country in the EU schedule is the Standard GSP, which offers reduced but not zero rates and which excludes most apparel categories from preference altogether. The arrangement that would substantially preserve our preference is GSP+, which is contingent on ratification and credible implementation of twenty-seven international conventions covering labour, environment, human rights, and good governance.

The United Kingdom's Developing Countries Trading Scheme, the successor to the UK's GSP, treats Bangladesh more generously than the EU's Standard GSP and continues to grant most apparel categories near-zero duty after graduation. The risk on the UK book, which is about a tenth of our exports, is therefore lower than on the EU book.

The TRIPS pharmaceutical waiver, which allows our generic drug industry to manufacture without holding the relevant patents on a number of essential medicines, runs on a separate schedule. The current LDC waiver, extended at the 2021 TRIPS Council, runs through end-2033 for LDCs in general; for graduating LDCs the transition arrangement is not yet fully settled, and Bangladesh has joined other LDCs in formally seeking an extension of pharma-sector treatment through 2029 to mirror the trade-preferences grace. The exports themselves are modest, around USD 245 million in FY2025, but the domestic supply story is far larger: roughly 95 percent of Bangladesh's pharmaceutical consumption is met by domestic manufacturers operating under the TRIPS waiver. The loss of that protection raises the per-unit cost of essential medicines for the domestic population in addition to closing the small export channel.

The concessional borrowing side is less dramatic but real. Graduation does not change our World Bank IDA classification immediately, because IDA classification is governed by income and creditworthiness criteria rather than LDC status. It does, however, change the eligibility for some development bank concessional windows and tightens the pricing on others. The marginal new external borrowing cost rises by something on the order of one to two percentage points over the next five years.

The fiscal channel adds a further layer. Customs and supplementary duties on imports currently account for roughly 22.3 percent of total tax revenue (FY2024 base: Tk 32,450 crore customs plus Tk 29,100 crore supplementary duty). As tariff bindings tighten with LDC graduation and MFN reciprocity norms apply more strictly, the revenue at risk ranges from Tk 6,155 crore (2.23 percent of total revenue) in the low scenario to Tk 11,079 crore (4.02 percent) in the mid scenario and Tk 15,388 crore (5.58 percent) in the high scenario.

These are the technical effects. The strategic effect is simpler: the implicit subsidies that have masked the under-collection of taxes and the under-development of higher-value exports for the last twenty years end on a known calendar date.

What 9.6 percent does to a knit shirt order

The competitive arithmetic is concrete. A knit cotton shirt landed in Rotterdam from Chittagong currently pays zero duty under EBA. The same shirt landed from Da Nang under the EU-Vietnam FTA pays zero duty. After November 2029, the Bangladesh shirt pays 9.6 percent under the EU Common External Tariff for chapter 61 apparel; the Vietnam shirt continues to pay zero.

For a shirt with a landed cost of USD 4 free-on-board Chittagong, that 9.6 percent is a duty incidence of about USD 0.38 per piece. On a one-million-piece annual order, the duty differential is USD 380,000. That is roughly the contribution margin on the order at the manufacturer level. The order does not stop being viable for the manufacturer; it stops being viable for the buyer, because the buyer's landed cost in the EU market is now 9.6 percent higher than the Vietnamese alternative for an identical garment.

Buyers reallocate sourcing in advance of the duty change, not after it. A buyer placing an order in May 2026 that will land in early 2030 is already making the post-graduation comparison. This is why the Mirpur manager's order book is already shrinking, with the explanation given as capacity reallocation and the actual reason as duty exposure.

The aggregate exposure is large. RMG accounted for roughly 84 percent of our total goods exports in FY2024. The EU plus the UK absorbed approximately 62 percent of our total exports, with most of that being RMG. A 9.6 percent (knit) to 12 percent (woven) duty differential on roughly USD 27 billion of EU-and-UK-bound RMG exports, applied across the post-2029 period, translates into a buyer-side landed-cost increase of approximately USD 2.6 to 3.2 billion annually. That cost is shared among the buyer, the consumer, and the supplier, depending on bargaining power. Bangladesh's bargaining power, as the dominant low-end supplier with limited substitution capacity at the manufacturer level, is on the supplier-share end of that spectrum. Half the duty incidence absorbing into manufacturer margins would compress the RMG sector's profit pool by something like USD 1.3 to 1.6 billion per year, or roughly 0.3 percent of GDP, at current export volumes.

This is before factoring in the secondary effects: the sourcing diversification that buyers will accelerate to Vietnam, Cambodia (which has now regained EU EBA-equivalent access after its 2020 partial withdrawal was eased), Indonesia, and India. The aggregate hit on RMG sector employment, currently around 4.5 million direct workers, could plausibly reach the low hundreds of thousands of jobs by the mid-2030s if no successor preference is secured and no domestic upgrading shifts the product mix toward less duty-sensitive segments.

The reserves position makes this story more urgent. Foreign exchange reserves collapsed from USD 46.2 billion at end-2021 to USD 21.4 billion at end-2024, less than four months of import cover. Remittances at USD 27.5 billion in 2024 stabilised the current account, but the cushion is thin. A meaningful contraction in the RMG export book at the same moment we lose concessional borrowing pricing is a recipe for an external sector crisis that the country has only just exited.

What the three-year grace window is actually for

The EU has been explicit that the three-year transition is intended for two things: (a) for the graduating country to negotiate a successor preference arrangement, most likely GSP+ in our case, and (b) for the graduating country's exporters to begin adjusting to a more competitive trade environment. The graceful interpretation of (a) is that we have until 2029 to put the GSP+ paperwork in order. The practical interpretation is that GSP+ designation typically requires twelve to eighteen months of EU institutional process after the applicant country has completed its ratification and implementation evidence. Counting backward from November 2029, the ratification work has to be substantially done by mid-2027. We are eighteen months from that point.

The blocker list is well known. Several core ILO conventions on freedom of association in EPZ and economic zone settings have been resisted by industry associations since the 2010s. The labour amendments required to give credible effect to the conventions even after formal ratification are politically charged inside the country. The Yunus transition government of 2024-2025 began the relevant institutional reviews; the Tarique Rahman BNP government, sworn in on February 17, has inherited the work and now has to either complete it or run out the clock.

The TRIPS work is more diplomatic than legislative. The current LDC waiver running through end-2032 gives Bangladesh a slightly longer breathing room than the trade preferences do, but the specific transition arrangement for graduating LDCs has been subject to several rounds of WTO discussion and is not fully settled. The right strategic posture is to lock in a clear, public TRIPS extension at the Twelfth or Thirteenth Ministerial Conference for Bangladesh-specific use, building on the precedent of the 2015 and 2021 extensions, rather than to assume that the 2032 horizon will hold automatically.

The bilateral track is the least developed and the most underutilised. Negotiations on a Bangladesh-Japan EPA have been on and off for more than a decade with no breakthrough. The Bangladesh-China FTA scoping has produced no operational text. There is no active Bangladesh-Korea FTA process. The GCC trade relationship is essentially energy-imports-for-remittance-exports rather than a goods-preference arrangement. Each of these bilateral tracks, if accelerated through the grace window, would diversify the EU exposure that currently dominates our preference dependence.

The cost of doing nothing is the buyer-side cost compounded over post-2029 years. The cost of doing the work is short-term political pain on the labour-conventions track and short-term diplomatic capital expenditure on the bilateral track. The arithmetic favours doing the work.

The diplomatic stack, in priority order

For the BNP government's first eighteen months in office, the LDC-cushioning agenda has a clear ordering.

First, EU GSP+ readiness. Ratify the seven outstanding ILO and human-rights conventions. Pass the credible implementation legislation, particularly the EPZ labour amendments that have been pending. File the formal GSP+ application by Q2 2027 at the latest, so the EU's institutional twelve-to-eighteen-month process completes by mid-2029. This is the single highest-yield diplomatic move available to us, because it preserves duty-free or near-duty-free access on the 51 percent of exports that go to the EU. It is also the move with the largest concentrated domestic political cost, on the BGMEA and EPZ industry lobby.

Second, the UK DCTS. The arrangement already preserves most preferences post-graduation, so the work here is diplomatic maintenance: ensure the UK is not pressured into harmonising downward with the EU, secure the duty-free coverage extension on knit and woven apparel categories that are administratively most exposed to reclassification, and lock in the rules-of-origin treatment that prevents cumulation from being used to chip away at the preference.

Third, TRIPS pharma. Push for a Bangladesh-specific TRIPS LDC-graduation transition extension, separately from the general LDC waiver, at the next WTO Ministerial. The argument is that our pharma industry serves a domestic population of 175 million whose access to essential medicines depends on this arrangement, and that the export-channel impact on developed-country IP holders is negligible. The asks are technically modest and politically achievable if the case is presented sharply rather than mixed into general LDC advocacy.

Fourth, the bilateral FTA pipeline. Bangladesh-Japan EPA negotiations should be concluded, not extended. Bangladesh-Korea, Bangladesh-GCC, and a serious revisit of Bangladesh-China should each be put on a defined eighteen-to-twenty-four-month negotiation schedule. The principle is diversification: each percentage point of export share moved out of the EU concentration is a percentage point less of exposure to the GSP+ negotiation outcome.

Fifth, the within-country adjustment. The RMG product mix has to shift up the value chain. Currently we are concentrated in the lowest-margin basic apparel categories where buyer substitution is easiest. Higher value-added segments (technical textiles, performance synthetics, branded private-label) carry larger margins that can absorb the post-2029 duty incidence without buyer flight. This shift requires industrial policy of a kind we have not seriously attempted: targeted finance for fabric mills (which we do not have at scale; we import most synthetic fabric), tax incentives tied to value-addition rather than to plant-level output, and skills upgrading inside the labour force. None of this is fast. All of it should have started a decade ago and can still meaningfully begin in 2026.

The diplomatic clock

The post-November-2026 calendar runs on three timetables. The graduation itself is November 24, 2026, six months from this morning's brief at the Brussels mission. The EU trade-preference grace window expires November 24, 2029, three and a half years from now. The TRIPS pharma waiver runs through end-2032 with the graduating-country status to be clarified. None of these dates moves because we wish them to.

The window for negotiation is narrower than the window for graduation. By Q2 2027 we need to have filed a credible GSP+ application. By end-2028 we need to have the bilateral FTA pipeline producing first signatures, not first MOUs. By end-2030 the within-country industrial policy adjustments should be showing first measurable effects on product mix. The BNP government has approximately twelve months from its February 17 inauguration to set this agenda and three years to execute the negotiating tracks on it.

The brief on the Brussels desk lists twenty-seven conventions. The costing on the Mirpur factory floor lists six percent of an order book lost in a single buyer cycle. Both are surface signals of the same underlying choice the country has not yet made: whether to use the grace window to negotiate the next twenty years of market access or to spend it confirming that we were not ready.

Sources

• WTO Ministerial Conference 2024 decision on preference extensions for graduating LDCs: wto.org (MC13, Abu Dhabi, February 2024)
• Bangladesh LDC graduation deferral request (Tarique Rahman government, CDP filing): bssnews.net/business/381418
• UN LDC Portal, Bangladesh graduation status: un.org/ldcportal/content/bangladesh-graduation-status
• LDC graduation trade and investment risk analysis: steinandpartners.com/investment-finance/bangladeshs-ldc-graduation-key-risks-recommended-actions/
• TRIPS pharma waiver and graduation risk for Bangladesh (1,000 drug registrations affected): tbsnews.net/bangladesh/1000-drug-registrations-stuck-bangladesh-risks-losing-trips-waiver-1224636
• Export destination shares FY2024 (EU 51.1%, US 18.5%, UK 10.5%, Canada 3.6%, Japan 2.7%): bdpolicylab data/bdpolicy.db, series 1523-1527, date 2024-06-30 (EPB)
• Customs revenue at risk scenarios (low Tk 6,155 cr / 2.23%, mid Tk 11,079 cr / 4.02%, high Tk 15,388 cr / 5.58%): BDPolicyLab fiscal model, precomputed customs-revenue risk scenarios (ch5)