The shape of the Bangladesh tariff: what 7,469 lines look like together
Most tariff guidance looks at one code at a time. This article does the opposite: it treats the whole Operative Tariff 2026-2027 as a single dataset and asks what the structure reveals.
Everything below is computed directly from the 7,469 lines published in the tariff schedule, reconciled against the National Board of Revenue's own online duty calculator. The figures are not estimates and not sampled — they are counts across the full set.
The headline numbers
Of 7,469 lines, 7,411 are ad-valorem — charged as a percentage of assessable value — and carry a computable Total Tax Incidence. The remaining 58 are specific-duty lines, charged by weight or quantity, where no meaningful percentage exists.
Across those 7,411 lines:
- Lowest TTI: 1%
- Highest TTI: 1,007.5%
- Mean TTI: 61.08%
A thousand-fold range. The arithmetic mean of 61.08% is close to the most common values, which is unusual for a distribution with such extreme outliers — and it tells you something: the extremes are rare, and most of the tariff is clustered.
The tariff is far more clustered than it looks
This is the most useful finding for anyone trying to hold the system in their head. Six values account for over 70% of all lines:
- 39.75% — 1,408 lines
- 64.25% — 1,356 lines
- 33.63% — 1,009 lines
- 28.73% — 576 lines
- 96.1% — 420 lines
- 80.18% — 327 lines
Those six figures cover 5,096 of 7,411 lines. If a calculated TTI lands on one of them, it is probably right. If it lands on an unusual value, it is worth a second look — not because odd values are invalid, but because they are rarer than you would expect and a transposed rate often produces one.
The clustering happens because the six component rates each take a small number of standard values, and most lines draw from the same handful of combinations. 64.25% — CD 25, SD 0, VAT 15, AIT 5, RD 5, AT 7.5 — is the tariff's default setting for an ordinary manufactured good, and it is the figure on smartphones as much as on bamboo seats.
Which component does the work
Looking at how each component is distributed shows what each is for.
VAT is a flat revenue instrument. 6,700 lines at 15%, 756 at 0%. It does almost no targeting.
Customs duty operates in five bands. 3,366 lines at 25%, 1,521 at 10%, 1,259 at 5%, 605 at 1%, 445 at 0%. Nearly half the tariff sits at the 25% ceiling.
Supplementary duty is the protective weapon. It is zero on 5,836 lines — 78% of the tariff. Where it appears it can be severe: 509 lines at 20%, 294 at 30%, 120 at 45%, 97 at 60%, 48 at 100%, 24 at 150%, 17 at 250%, 34 at 350% and 12 at 500%. SD is not a revenue instrument; it is a policy statement about specific goods.
Regulatory duty is the adjustable one. 4,363 lines at 0%, 2,855 at 5%, and small groups at 10, 15, 20 and 25%.
The practical lesson, which repeats throughout the tariff: customs duty tells you remarkably little about what you will pay. In chapter 33 every single line carries 25% CD, and the TTI still ranges from 64.25% to 189.2%.
The extremes
The most taxed goods in Bangladesh are cars. Four lines reach 1,007.5% — large-engine motor cars, both reconditioned and new, CBU. Six more sit at 860.5%, including CKD cars and double-cabin pickups. Supplementary duty of up to 500% drives all of it. See motorcycle or car.
The least taxed are medical and agricultural. At 1%: invalid carriages (87131000 and 87139000), angiographic catheters (90189030), artificial body parts (90213900) and hearing aids (90214000). At 2%: seed potatoes, dried peas, chickpeas, lentils and durum wheat seed.
The policy reads clearly off the data without any need for commentary: mobility aids and food security at one end, private motoring at the other.
Chapters ranked
Averaging TTI by chapter gives a rough map of where the burden sits.
Heaviest: chapter 22, beverages and spirits, at 457.39% across 24 lines — by a wide margin the most taxed chapter in the tariff. Then chapter 24, tobacco, at 331.01%. Then arms (chapter 93) at 142.06%, vehicles (chapter 87) at 135.66% across 510 lines, cosmetics at 122.71% and footwear (chapter 64) at 113.89%.
Lightest: chapter 01, live animals, at 12.75%. Then fertilisers at 14.63%, pharmaceuticals at 17.95%, vegetable saps at 18.00%, animal fodder at 19.96% and raw hides at 21.38%.
The pattern is consistent and deliberate. Agricultural inputs, pharmaceuticals and industrial raw materials are protected. Alcohol, tobacco, vehicles and discretionary consumer goods carry the load.
What this means when you look up a code
Three practical takeaways.
Expect a clustered value. Most lines land on one of a dozen or so figures. An unfamiliar result is worth re-checking against the breakdown on the page.
Read SD before CD. If supplementary duty is zero — as it is on 78% of lines — your TTI will be moderate almost regardless of the other rates. If it is not zero, it will dominate everything else.
Know which side of the protective line your product sits on. The tariff is built to tax finished consumer goods and admit inputs. Where your product falls on that gradient predicts the rate better than its chapter does.
A note on method
These figures were produced by querying the complete tariff dataset published on this site. The Total Tax Incidence for each line is computed from the six published rates using the cascade below, which was tested against all 7,411 ad-valorem lines and reproduces NBR's published figure on every one:
CD = AV x cd% RD = AV x rd% SD = (AV + CD + RD) x sd% VAT = (AV + CD + RD + SD) x vat% AT = (AV + CD + RD + SD) x at% AIT = AV x ait% TTI = CD + RD + SD + VAT + AT + AIT
Chapter averages are unweighted means of the ad-valorem lines in each chapter — they describe the schedule, not actual trade, since a chapter's 500 rarely-used lines count as much as its two high-volume ones. Read them as a map of policy intent rather than of revenue.
The underlying data is the Operative Tariff 2026-2027 as published by the National Board of Revenue. You can verify any figure here against the individual product pages, or browse the whole structure from the chapter index.