Bangladesh's average bound tariff at the WTO is 159%. It applied 13% in 2021. It is far from the only economy with that much room.

When a country joins the World Trade Organization, it files a schedule of concessions that specifies a maximum tariff rate — a bound rate — for the product categories it agrees to bind. A bound line is a legal ceiling, enforceable at the WTO; a line left unbound carries no ceiling. The rate a country actually charges on a given day is the applied rate, which on a bound line can sit anywhere at or below that ceiling. The gap between them is known as binding overhang, or tariff water: unused room to raise tariffs within the rules. Measured here as the gap between a country's average bound and average applied rate — the World Bank figures, which count unbound lines as zero — it is an aggregate proxy; the binding limit on any single product is that product's own line. The World Bank compiles both rates annually from UNCTAD TRAINS. This piece pulls the two World Bank series for 2021, chosen for its bound-rate coverage, and computes the gap for every economy with both figures. That yields 140 economies with matched data; Switzerland is set aside as a data artifact (see below), leaving a 139-economy analytical panel. Both series use the simple-mean method — an unweighted average across all tariff lines — so the comparison is methodologically aligned, though the specific tariff lines each series covers may differ. Bangladesh holds the largest gap among the 139 economies. Its WTO-bound rate averages 158.9% across all product lines. Its applied rate is 12.6%. The aggregate gap is 146.3 percentage points. Nigeria and Tanzania follow: bound at 121.3% and 120.0% respectively, applied at 12.7% and 12.0%, with gaps of 108.6 and 108.0 points. Zambia, Kuwait, the Gambia, Mozambique, Mauritius, Myanmar and Kenya round out the top ten, each carrying more than 80 points of unused tariff space. At the other end of the distribution, the United States has a bound rate of 3.6% and an applied rate of 2.8% — a gap of 0.8 points. The European Union (which negotiates as a single WTO member) runs a bound rate of 4.4% and an applied rate of 2.2%, a gap of 2.2 points. Canada and Japan sit in the same narrow band: gaps of 2.9 and 0.9 points respectively. Australia is an outlier among rich economies — bound at 9.6%, applied at 1.9%, a gap of 7.7 points. India, a frequent subject of WTO trade disputes, carries a bound rate of 52.0% against an applied rate of 9.9%: 42.1 points of unused room. Indonesia's gap is 31.3 points (bound 37.3%, applied 6.0%). Brazil's is 18.3 points (bound 31.5%, applied 13.2%). China's is narrow: 4.7 points (bound 10.0%, applied 5.3%). Across the 139-economy panel, the median gap is 12.6 percentage points. Fifty-seven of 139 economies have gaps exceeding 20 points; 82 of 139 exceed 10 points; 45 of 139 show gaps below 5 points — and most of those are wealthy economies or members of customs unions where bound rates are already low. Two economies, Hong Kong and Macao, carry a bound rate of 0.0% and apply 0.0% — the only cases where the gap is precisely zero. Two economies show applied rates nominally above their bound rates in the 139-economy panel: Norway by 0.7 points and Cote d'Ivoire by 0.9 points. These are not confirmed WTO violations. At the aggregate simple-mean level, rounding in bound-rate schedules and differences in the tariff lines covered by each series can produce small apparent inversions. Confirming a WTO violation requires a product-line analysis, which this aggregate series does not support. Switzerland reports a bound rate of 0.0% in the World Bank series across all available years, producing a large apparent inversion against its 4.1% applied rate. The World Bank indicator records Switzerland's simple-mean bound rate as 0 consistently, suggesting a data artifact — Switzerland does have WTO commitments — and Switzerland is excluded from the panel analysis on this basis.

The following is analysis, not fact. At the tariff-line level, binding overhang directly measures unused headroom; the country-average gap used here is an aggregate proxy for how much room the rules-based trading system leaves unused. On the average figures, a country's committed ceiling can sit five, ten, or twenty times above its current average rate — a commitment to the status quo only as long as the status quo suits the country, though on any individual product the binding line is reached first ^mmodeled. The WTO's dispute mechanism, the schedule of concessions, and the phrase 'rules-based order' all treat the bound rate as the system's anchor. But for 57 of the 139 economies with comparable data, that anchor sits more than 20 percentage points above the average applied rate. The largest observed gaps are concentrated among developing economies ^mmodeled. They are predominantly low- and lower-middle-income states that joined the WTO (or its predecessor, GATT) in rounds when developing countries were not pushed toward aggressive binding commitments ^mmodeled. Bangladesh's 159% average bound rate reflects a negotiating outcome, not an intent to charge 159% ^mmodeled: the country agreed to bind its tariffs — itself a concession — without being pushed to bind them low. The ceiling is the price of membership, the applied rate is the price of imports, and the gap between them is the residue of each round of negotiations ^mmodeled. That residue matters now in a way it did not in the 1990s or 2000s ^sspeculative. The 2025 wave of tariff escalations by major economies — the United States in particular — has focused attention on who can retaliate and how. Applied rates can be raised toward the bound level without any WTO procedure. On the average figures, a country with 80 or 100 points of tariff water could raise applied rates substantially within the rules, though line-by-line ceilings would bind before the average gap is exhausted ^mmodeled. Countries with 1 or 2 points of average water — the US, EU, Japan — have little equivalent room: their bound and applied rates are close, which may be one reason negotiators in Washington and Brussels have acted through emergency safeguard clauses and Section 232 rather than simply dialing up applied rates ^mmodeled. The asymmetry in tariff water may, in part, translate into an asymmetry in trade-war ammunition ^sspeculative. The gap also bears on investment planning. A large country-average overhang may indicate greater tariff-policy uncertainty for some manufacturers weighing where to invest, depending on the relevant product-line bindings ^mmodeled. The WTO schedule legally commits a member to hold rates at or below the per-line ceilings; it does not guarantee they will remain near the floor. Predictability at the ceiling is not the same as predictability at the applied rate ^sspeculative.

The strongest counter to the tariff-water framing is that the gap is not a latent threat but a design feature that has functioned as intended for three decades. The GATT and WTO rounds deliberately allowed developing countries to set high bound rates as a development prerogative — the ceiling was always understood to be a safety valve, not a target. Historically documented episodes of significant applied-rate increases exploiting large tariff water appear uncommon outside genuine economic crises (Argentina in 2001-02 among them). The existence of the gap does not predict its use, and treating it as 'unexploded ammunition' attributes intent to an accounting identity ^sspeculative. A second counter is methodological. The simple-mean bound rate — an unweighted average across all tariff lines, including unbound lines counted as zero — is an imperfect proxy for actual legal headroom. Bangladesh's 159% figure averages across lines that include some very high agricultural bound rates and many lower industrial rates; the headline number likely overstates the room on most goods and understates the political constraint on raising rates on food, where consumer prices are politically sensitive ^mmodeled. Product-line-level analysis tells a different story than the country average. A third counter applies specifically to the low-gap economies. The US, EU, and Japan have low gaps not because the WTO forced them there but because they chose, through eight rounds of negotiations, to liberalize deeply. The constraint they face is one they imposed on themselves as part of extracting concessions from others. From that vantage, the asymmetry is the reward for having led liberalization, not a vulnerability.

From India's vantage, the 42-point gap reads as a policy instrument, not a risk. On this view — the one a middle-income WTO member might well take — the ability to raise applied rates without WTO violation is a development tool that economies at India's income level retained in their Uruguay Round schedules. Where a critic sees policy uncertainty, this vantage sees policy space. Both readings describe the same number: the gap reads as a risk from the exporter's side of the ledger and as a right from the importer's side.