The State of Bangladesh Public Finance

Bangladesh collects about 7.5% of GDP in taxes (NBR/CEIC, Dec 2024; World Bank WDI 7.4-7.7% range), one of the lowest ratios in Asia and stuck in the 7-8% band for years. That base is too thin to fund the public investment the country still needs, and the shortfall is administrative, not legislative: only 1.2 million people pay income tax against 10.8 million TIN holders (NBR FY2023-24), and direct tax expenditure runs at 2.39% of GDP, BDT 107,132 crore, equal to 99% of direct tax collection (NBR Direct Tax Expenditure Report FY2022-23). On 1 June 2026 the UN Committee for Development Policy recommended deferring LDC graduation by three years to 24 November 2029 (ERD; CDP), so Bangladesh now has a defined window, not a cliff. The governing thought: use the three deferral years to close the TIN-to-payer funnel and prune exemptions, because those two administrative moves, not new tax law, are what determine whether the wider tax base exists when preferences finally expire.

On 1 June 2026 the UN Committee for Development Policy recommended deferring Bangladesh's LDC graduation by three years, from 24 November 2026 to 24 November 2029 (ERD; CDP), pending ECOSOC review on 21-22 July and final UN General Assembly approval expected in September 2026. That converts a deadline into a window. The question is no longer whether Bangladesh is ready, it is whether the next three years are spent fixing the revenue base or watching it stay flat. Bangladesh collects about 7.5% of GDP in taxes (NBR/CEIC Dec 2024; World Bank WDI 7.4-7.7% range), roughly 7.5pp below the 15% benchmark and stuck in the 7-8% band for years. The shortfall is administrative, not legislative: 10.8 million people hold a Tax Identification Number, but only 1.2 million actually pay income tax (NBR FY2023-24). Direct tax expenditure foregone through exemptions runs at 2.39% of GDP, BDT 107,132 crore, equal to 99% of direct tax collection (NBR Direct Tax Expenditure Report FY2022-23). Closing the registration-to-payment funnel and pruning that exemption regime would recover revenue worth a large multiple of any plausible rate increase, and both can begin under existing law.

A thin base, not a cyclical dip

At about 7.5% of GDP, Bangladesh's tax take is among the lowest in Asia and well below the 15% benchmark used for the FiscalAnalysis tax-effort gap. The Ministry of Finance Monthly Fiscal Report puts actual NBR collection at 7.32-7.38% of GDP in FY24, and CPD records the ratio sliding to 6.78% in FY25, its weakest reading in the recent series. This is structural, not a one-year dip. Composition compounds the weakness. VAT supplies 38% of tax revenue, income tax 32%, and customs 22% (NBR composition). Customs is the most exposed line: once graduation takes effect in 2029 it deepens trade liberalisation and erodes the tariff base, so the 22% slice is structurally shrinking precisely when revenue is most needed. Leaning harder on VAT and income tax is therefore not optional, it is the only direction the arithmetic allows.

The funnel, not the rate, is the binding constraint

The base crisis reduces to one funnel. Of 10.8 million TIN holders, 4.4 million file a return, and only 1.2 million pay any income tax (NBR FY2023-24). The collapse from registration to payment is administrative. NBR cannot cross-reference income, asset, and spending data at scale, so filers report zero liability through exemptions, blanket deductions, or underreporting that goes undetected. CPD records the identical pattern historically, with 7.5 lakh payers against 22 lakh TIN holders. The implication is direct: the gap between 10.8 million registered and 1.2 million paying is a collection failure, and collection failures are closed with data infrastructure and compliance design, not with higher rates that the unregistered never see.

Spending is capped by revenue, and even that cap is unspent

Government expenditure is low not by consolidation design but because thin revenue imposes a hard ceiling, then the state fails to spend even within it. ADP utilisation fell to 67.85% in FY2023-24 (IMED), the weakest in two decades, and ADP spending contracted a further 25.5% in FY25 (World Bank). Procurement delays, land-acquisition bottlenecks, and weak project management drive the shortfall. The result is a double penalty: the development envelope is small because revenue is small, and the small envelope is then under-executed.

Debt makes the squeeze tighter on schedule. As megaproject grace periods expire, interest becomes a first charge on a revenue base that has not grown to meet it, and graduation in 2029 narrows the concessional finance that has cushioned the transition so far (ERD; CPD).

Recommendations

1. NBR closes the funnel through mandatory e-filing and NID-TIN-bank linkage by FY2027.

Require electronic filing for all 10.8 million TIN holders and integrate NID and bank-account data so income can be cross-referenced automatically. Owner: NBR, with a Finance Division data-sharing mandate. Target: lift the 1.2 million actual payers toward the 4.4 million who already file by FY2029, raising collection without touching statutory rates. Success signal: the actual-payer count and the payer-to-filer ratio reported in NBR's annual return, tracked each fiscal year against the 2026 baseline.

**2. NBR and the Finance Division prune the direct tax expenditure with a published sunset schedule.** Publish a sunset schedule for exemptions, deductions, and preferential rates, retiring the lowest-justified first. Owner: NBR and Finance Division. The pool is BDT 107,132 crore, 2.39% of GDP (NBR FY2022-23); recovering even one-third adds revenue worth roughly 0.8pp of GDP, the single largest pool reachable under existing law. Success signal: a year-on-year fall in reported direct tax expenditure as a share of GDP in the next Tax Expenditure Report.

3. Enforce tax-linked access to public services.

Make proof of return submission a condition for trade licences, vehicle registration, and bank loans, as the 2024 White Paper on the Bangladesh Economy recommends. Owner: NBR with the relevant licensing authorities. This converts the dormant TIN base into filers at near-zero administrative cost by giving registration consequences. Success signal: the return-filer count moving toward the 10.8 million registered base.

4. Reweight collection toward VAT and income tax ahead of customs erosion.

With customs (22% of tax revenue) structurally shrinking after 2029 graduation, the Finance Division should set explicit FY2027-FY2029 revenue-share targets that shift the burden onto VAT and income tax. Owner: Finance Division. Success signal: customs falling as a share of tax revenue while the tax/GDP ratio holds or rises, not falls, through the deferral window.

What would change this view

If the headline tax/GDP figure is revised up by a higher FY2026 outturn, or if NBR data confirms the actual-payer count has already risen materially above 1.2 million, the "collection failure" framing weakens and the case shifts toward rate and base-rate reform.

The strongest counterargument is sequencing: with graduation now recommended for 2029 rather than 2026, the pressure to move fast eases, and the Finance Division may prefer to bank the deferral and defer the politically costly exemption cuts. That is the trap. The deferral is a preparation window the CDP explicitly conditioned on reform progress, not a postponement of reform; if the three years pass with the funnel and the exemptions intact, graduation in 2029 lands on the same 7.5% base and the concessional-finance loss bites just as hard, only later.

Sources: IMF Article IV Consultation 2025 (Country Report No. 26/24); World Bank WDI; National Board of Revenue; Ministry of Finance Monthly Fiscal Report; CPD; Economic Relations Division; UN Committee for Development Policy; IMED; White Paper on the Bangladesh Economy 2024.

Tax revenue, fiscal balance, and public debt benchmarks from IMF Article IV Consultation Bangladesh 2025 (IMF Country Report No. 26/24, Board concluded 26 January 2026). Headline tax/GDP anchored at 7.5% (NBR/CEIC Dec 2024; World Bank WDI 7.4-7.7% range), reconciled against the Ministry of Finance Monthly Fiscal Report FY24 (7.32-7.38%) and CPD FY25 (6.78%). Historical tax/GDP trajectory (2001-2021) sourced from data lake (source=world_bank, series_id=60, wb_tax_revenue_pct_gdp, WDI GC.TAX.TOTL.GD.ZS; lake snapshot 2026-05-12). Government expenditure and debt series from World Bank WDI (GC.XPN.TOTL.GD.ZS, DT.DOD.DECT.GN.ZS). Revenue series from WDI (GC.REV.TOTL.GD.ZS). NBR revenue collection from NBR monthly press releases. Tax expenditure from NBR Direct Tax Expenditure Report FY2022-23. Series extracted via FiscalAnalysis analyzer.