Executive judgment
The WTO, International Trade Centre and UN Trade and Development have released World Tariff Profiles 2026, a 234-page statistical account of tariffs and selected non-tariff measures across more than 150 economies and customs territories.1 Its central lesson is less obvious than “some countries have high tariffs.” The report shows that a credible trade strategy must distinguish between the tariff a government is legally entitled to charge, the tariff it normally applies, the tariff actually paid on a given consignment, and the additional legal instruments that may neutralise the apparent opening.
This matters especially for a rules-based but development-conscious trading order. Average tariffs are useful indicators, but they can conceal tariff peaks in politically sensitive sectors, complex non-ad-valorem duties, unbound lines, preference-dependent access and trade-remedy exposure. Conversely, a high simple average does not establish that every import faces a high effective burden: trade-weighted rates may be lower because trade is concentrated in lower-duty lines, because preferences are used, or because firms have already reorganised supply chains around protection. Neither reading is enough on its own.
Zwarte Peper’s position: the appropriate response to uneven market access is not rhetorical protectionism or regulatory imitation. It is disciplined reciprocity: publish the real barriers, distinguish legitimate regulation from disguised exclusion, reduce domestic input costs, and negotiate rules that make access usable for firms from every region.
India is a particularly important illustration. The 2026 profile reports a 2025 simple average MFN applied tariff of 16.0% overall, 36.4% for agricultural goods and 12.8% for non-agricultural goods; the corresponding trade-weighted averages are 7.9%, 24.4% and 6.9%.2 These figures explain why India can simultaneously look highly protected in a line-by-line negotiation and commercially accessible in particular industrial supply chains. They also explain why the EU–India FTA should be evaluated as an implementation project, not simply a political announcement: its value depends on product coverage, origin qualification, staging, standards and customs practice.
1. How to read the report without drawing the wrong conclusion
World Tariff Profiles standardises national schedules at the six-digit Harmonized System level. It provides simple and trade-weighted averages, duty-free shares, tariff peaks, non-ad-valorem duties, binding coverage, the number of distinct rates and a measure of dispersion. It also reports selected anti-dumping, countervailing and safeguard measures. This is unusually valuable because it makes cross-country comparisons more disciplined; it is not a substitute for transaction-level customs advice.
| Concept | What it measures | Why it can mislead if used alone |
|---|---|---|
| Bound tariff | The ceiling committed in the WTO schedule, subject to the coverage of that schedule. | It is a legal maximum, not necessarily the rate paid today. |
| MFN applied tariff | The ordinary rate applied to WTO partners absent a preference or another special regime. | It excludes preference utilisation, exemptions, trade remedies and domestic charges. |
| Simple average | The average across tariff lines or pre-aggregated HS-6 subheadings. | A niche line and a major import line receive equal weight. |
| Trade-weighted average | The average weighted by observed import value. | Trade may already have been diverted away from high-duty lines, making protection look lower. |
The report itself flags three cautions. First, a comparison between bound and applied rates is only fully meaningful where binding coverage is complete. Second, some commitments may not be fully implemented. Third, converting specific, compound or mixed duties into ad-valorem equivalents requires unit-value assumptions. Those are not footnotes for specialists; they are conditions for responsible public claims about who is “open” or “closed.”3
Evidence boundary: a profile is not a live customs ruling
The 2026 edition combines different observation periods: the headline MFN and binding indicators in this article are 2025 data; the India import figures are 2024 trade values; and the remedies tables record positions at 31 December 2025. That makes the publication a disciplined comparative baseline, not confirmation of the duty payable on a shipment today. A current tariff determination still requires the precise tariff classification, origin, valuation, preference evidence, quota position, applicable trade-remedy notices and the law in force on the date of import.3
There is a further commercial caution. The profile is predominantly MFN-based. A Dutch exporter using an FTA, a least-developed-country preference, a tariff-rate quota, a drawback programme or a sector-specific exemption may face a materially different rate. The right sequence is therefore: identify the product and origin; identify the ordinary rate; test preference eligibility; then model the complete landed cost, including non-tariff compliance and remedy risk.
2. What the data says about the real geography of protection
A consistent comparison test. Before attaching a political label to any tariff regime, compare the same variables across jurisdictions: the trade-weighted and simple average; tariff peaks in sensitive lines; preference availability; binding coverage; non-ad-valorem complexity; and the use of trade remedies. This is the minimum discipline needed to distinguish development policy, sectoral protection and genuinely usable market access.
The report does not support a simple developed-versus-developing-country story. The European Union’s 2025 MFN simple average is 4.9% overall and 4.7% for non-agricultural goods, but 29.2% for agricultural goods; its trade-weighted agricultural average is 29.0%.4 The United States records a low overall MFN simple average of 3.3%, while agricultural protection and contingency measures remain commercially material in particular sectors. India’s wider gap between agricultural and non-agricultural protection reflects a different political-economy settlement: food security, farm livelihoods, industrial policy and fiscal autonomy carry more weight at the border.
That does not make each system equally defensible. It means a credible critique must apply the same test everywhere. When a wealthy jurisdiction presents itself as a champion of openness while maintaining very high farm tariffs, complex quota administration or aggressive contingency protection, it should be assessed by the market access it provides in the affected lines, not its economy-wide average. When a developing economy uses broad tariffs to manage structural adjustment, it should account for the burden placed on its own downstream producers, consumers and exporters. Rules-based trade requires symmetry of analysis before it can achieve symmetry of obligation.
India: a large preference margin, but not a blank cheque
India’s 2025 profile shows 74.3% binding coverage and a 50.8% simple average final bound tariff overall, against a 16.0% simple average MFN applied tariff. Agricultural applied tariffs are much higher than non-agricultural ones, and 71.5% of agricultural HS-6 subheadings carry duties above 15%, compared with 32.6% of non-agricultural subheadings.2 The spread between the bound and applied average is often called “binding overhang.” It creates policy space to raise applied rates without breaching the ceiling. For investors, that can add a form of legal uncertainty even where today’s commercial tariff is manageable.
For Dutch and European firms, the important conclusion is not that India is an impossible market. India imported US$720.0 billion of goods in 2024, including US$679.6 billion of non-agricultural goods. Its trade-weighted non-agricultural MFN average of 6.9% is substantially below its simple average, indicating that the composition of actual imports matters greatly.2 But the commercial case must be made by sector. Tariffs remain only one layer alongside standards, licensing, local-content criteria, data rules, logistics and payment risk.
Board implication
Do not use a country average to approve or reject a market. Require a tariff-and-regulation matrix at HS-code and business-model level, with an explicit distinction between legal ceiling, current MFN rate, likely preferential rate, non-tariff requirements and trade-remedy exposure.
3. Trade remedies are no longer peripheral
The most commercially important part of the publication may be its reminder that anti-dumping, countervailing and safeguard measures operate alongside ordinary tariffs. These instruments are lawful only under defined WTO disciplines, but their economic effect can be immediate: a product that appears to have a low ordinary rate may become uncompetitive when a final measure applies. Trade counsel should therefore treat remedies as an upfront market-entry variable, not a dispute that can be addressed after the first shipment.
India initiated 81 anti-dumping investigations in 2025, implemented 42 final measures and had 175 measures in effect at year-end, covering 504 HS-6 product categories. As an exporter, it faced 111 final anti-dumping measures covering 232 HS-6 categories.5 The European Union initiated 30 investigations, had 144 measures in effect and faced 28 measures as an exporter. The figures do not prove abuse by either side. They prove that the commercial environment is shaped by contingent protection as well as ordinary tariff schedules.
The sectoral concentration is equally important. At 31 December 2025, the report records 1,031 anti-dumping measures in minerals and metals and 527 in chemicals, far more than in most consumer-goods categories. Flat-rolled steel products, tubes, pipes, certain plastics and tyres feature prominently among the most frequently affected headings.6 This is precisely where industrial policy, energy cost, subsidies, overcapacity allegations and strategic-supply-chain politics collide.
There are two errors to avoid. The first is to call every remedy protectionist. Subsidisation and injurious dumping can distort competition, and the WTO framework provides a process to respond. The second is to treat formal legality as an adequate commercial safeguard. Investigations impose legal cost, uncertainty and customer hesitation before a final duty is imposed. For capital-intensive exporters, the risk must be priced into contracts, inventory and financing.
4. What the report changes in the EU–India FTA conversation
The WTO data strengthens, rather than replaces, the case for a high-quality EU–India agreement. The largest gains are likely to come where the agreement converts meaningful Indian MFN rates into a usable preference and removes an associated administrative uncertainty. The report shows why headlines such as “over 90% of tariff lines” are incomplete. The decisive questions are which lines account for trade and investment, how quickly rates stage down, whether a company can demonstrate origin, and whether non-tariff administration changes in practice.
| Question | Evidence required | Decision consequence |
|---|---|---|
| Is the preference valuable? | Product-level MFN rate, negotiated staging and projected import value. | Separates material savings from symbolic liberalisation. |
| Can it be claimed? | Product-specific origin rule, bill of materials, supplier declarations and certification process. | Determines whether the nominal preference survives operational reality. |
| Can a remedy override it? | Current measures, open investigations, historic cases and supplier-country exposure. | Changes price, contract and inventory assumptions. |
| Will non-tariff friction remain? | Standards, testing, registration, local content, licensing, customs release and data requirements. | Identifies the true critical path to market. |
The report also gives a more balanced account of why India has negotiated selectively. An economy with higher agricultural protection and considerable binding overhang has more to exchange in a trade negotiation, but it also faces domestic distributional constraints. A sustainable agreement must give India credible gains in services, mobility, technology, investment and its own exports; it cannot be designed as a one-way extraction of goods-market access. That is not a concession to protectionism. It is the practical condition for a durable rules-based bargain.
5. The wider lesson: open trade needs institutional equality
Criticism of Western bias in global trade should be evidence-led. The relevant claim is not that advanced economies should have no regulatory standards or that all tariffs are equivalent. It is that the distribution of rule-making power, compliance capacity and adjustment costs matters. A system that treats its own farm protection, subsidy architecture, carbon-adjacent requirements or technical standards as neutral while describing developing-country policy space as an aberration will lack legitimacy and eventually lose compliance.
The alternative is neither autarky nor a race to the bottom. It is a system of published rules that distinguishes a justified public measure from a disguised restriction; accepts equivalent outcomes rather than demanding institutional imitation; gives lower-capacity economies real access to technology, finance and participation; and disciplines every major economy’s use of unilateral leverage. Trade law should constrain arbitrary power at the border wherever it is exercised.
This approach is also commercially intelligent. Supply chains are being diversified across South Asia, Southeast Asia, Africa, Latin America and the Middle East. Firms need standards information, digital customs systems, trusted payment channels, logistics and legal remedies that work across those routes. The winning jurisdictions will not be those that merely raise or lower an average tariff. They will be those that make entry predictable, evidence-based and affordable for new competitors.
6. Practical actions
For public institutions
- Publish a machine-readable tariff, remedy and regulatory-access dashboard, including the difference between bound, MFN and preferential treatment.
- Review tariff peaks as part of domestic productivity policy: test the cost imposed on downstream users, exporters and consumers.
- Build review mechanisms for standards and trade remedies with reasons, time limits and accessible challenge routes.
- Use mutual recognition, equivalence and international standards where they deliver the policy outcome at lower fixed cost.
For companies and advisers
- Replace country-average assumptions with an HS-code-level market-access file.
- Screen anti-dumping, countervailing and safeguard exposure before quoting or locating inventory.
- Model the FTA preference only after an origin and data-evidence review.
- Allocate tariff, remedy and regulatory-change risk expressly in distribution, supply and project-finance contracts.
The report’s real value is not a ranking of countries. It is a discipline of attention. The border is a composite institution. Good trade policy makes it legible and contestable; good commercial strategy reads every layer before capital is committed.
Authorities and methodology
- WTO, ITC and UN Trade and Development, World Tariff Profiles 2026, pp. 1–3, 6–7 and 201–219. The report uses national tariff schedules and standardises statistics at the HS six-digit level.
- World Tariff Profiles 2026, India tariff profile, p. 104.
- World Tariff Profiles 2026, general note and technical notes, pp. 2 and 7.
- World Tariff Profiles 2026, summary tables, pp. 8–9 and 20–21 (European Union and United States); figures are MFN tariff indicators for 2025.
- World Tariff Profiles 2026, Table 2.1.1, pp. 202–204.
- World Tariff Profiles 2026, Tables 2.2.1 and 2.3.1, pp. 215–217.
Editorial note. This article analyses a statistical publication released in 2026. Tariff treatment remains product-, origin-, valuation- and date-specific. It is general research, not customs, trade-remedy or legal advice.