Written and maintained by CASRAI Editorial Board
Published
On this page
Trade policy encompasses the statutory tariffs, non-tariff measures, trade remedies, and international agreements through which national governments regulate import and export commerce, evaluated in quantitative economics through tariff restrictiveness indices, value-added protection formulas, and econometric gravity models. In empirical research and policy analysis, quantifying trade policy requires calculating nominal and effective rates of protection across supply chains and estimating bilateral trade elasticities while rigorously controlling for multilateral resistance and supply-chain linkages.
Definition and Quantitative Foundations of Trade Policy
In quantitative economic analysis, trade policy refers to the complete set of price-based, quantity-based, and regulatory interventions implemented by sovereign authorities to alter the terms, volume, composition, or geographical origin of cross-border trade flows. While traditional policy discussions often treat trade policy as a qualitative stance between free trade and protectionism, empirical trade economists and policy analysts rely on econometric modeling and empirical methods to quantify the exact economic distortions, welfare transfers, and supply-chain impacts caused by specific policy instruments.
The primary instruments of trade policy fall into two broad structural classes:
- Price-based measures: Customs tariffs, which include ad valorem tariffs (a fixed percentage levy on the customs value of imported goods), specific duties (a fixed monetary fee per physical unit or volume, such as dollars per metric ton), and compound duties combining both mechanisms. Price measures also encompass export subsidies, export taxes, and trade-contingent border adjustments such as anti-dumping duties and countervailing duties.
- Non-tariff measures (NTMs): Policy interventions other than ordinary customs tariffs that economically influence international trade in goods. Under the international classification developed by the Multi-Agency Support Team (MAST) and UNCTAD, NTMs include Sanitary and Phytosanitary (SPS) measures, Technical Barriers to Trade (TBT), pre-shipment inspections, tariff-rate quotas (TRQs), export restrictions, and rules of origin attached to preferential trade arrangements.
A crucial distinction in international trade law and empirical analysis is the divergence between legal commitments and applied policy. Under the General Agreement on Tariffs and Trade (GATT 1994) administered by the World Trade Organization (WTO), member countries commit to ceiling rates known as bound tariffs (under GATT Article II). However, governments frequently apply lower rates, termed Most-Favoured-Nation (MFN) applied tariffs (governed by GATT Article I). The mathematical difference between a country’s bound tariff ceiling and its applied MFN rate is defined as the binding overhang (or tariff water). In econometric policy modeling, a substantial binding overhang represents regulatory policy uncertainty: even without violating international treaties, an administration can unilaterally raise applied tariffs up to the bound ceiling, which empirically dampens foreign direct investment and long-term trade commitments.
Mathematical Formulas for Measuring Trade Policy Protection
To rigorously evaluate the economic incidence of trade policy, empirical researchers use mathematical indices that capture trade distortions across supply chains. The two primary mathematical frameworks used by the World Trade Organization, the World Bank, and academic economists are the Effective Rate of Protection (ERP) and the Structural Gravity Equation.
The Effective Rate of Protection (ERP) Formula
Nominal tariffs measure the price increase on the gross output of an imported product. However, modern production processes rely on imported intermediate inputs, components, and raw materials. If a domestic industry imports components subject to tariffs, the protection afforded to the domestic manufacturing process depends on both the tariff on the final good and the tariffs levied on its required inputs. The Effective Rate of Protection (ERP), formulated by Max Corden (1966) and Bela Balassa (1965), measures the percentage increase in domestic value added per unit of output permitted by the entire tariff structure compared to a free-trade benchmark:
ERPj = (V′j − Vj) / Vj = (tj − ∑i aij ti) / (1 − ∑i aij)
Where every mathematical symbol is defined as follows:
- ERPj: The effective rate of protection for domestic industry or final output j, expressed as a decimal or percentage change in domestic value added per unit.
- tj: The nominal ad valorem tariff rate levied on the imported final commodity j (where 0.15 denotes 15%).
- ti: The nominal ad valorem tariff rate levied on imported intermediate input i.
- aij: The cost share of intermediate input i in the total production value of final commodity j under free-trade conditions (evaluated at world prices). The condition ∑i aij < 1 must hold for any viable production process.
- Vj: The domestic value added per unit of output under undistorted free trade, defined mathematically as Vj = 1 − ∑i aij.
- V′j: The domestic value added per unit of output under the prevailing tariff structure, defined as V′j = (1 + tj) − ∑i aij (1 + ti).
The mathematical properties of the ERP equation demonstrate three fundamental trade policy states:
- Tariff Escalation (ERPj > tj): When the nominal tariff on the finished product exceeds the weighted average tariff on intermediate inputs (tj > ∑i aij ti), the effective protection granted to domestic processing value added is strictly greater than the nominal final tariff.
- Negative Effective Protection (ERPj < 0): When intermediate input tariffs exceed final product tariffs such that ∑i aij ti > tj, the domestic value added with tariffs is smaller than under free trade. Domestic manufacturers face an effective tax on domestic processing, putting them at a structural disadvantage against foreign producers who import finished goods.
- Uniform Tariff Protection (ERPj = tj): When identical ad valorem tariffs are applied uniformly across all intermediate inputs and finished outputs (tj = ti for all i), the ERP collapses to the nominal tariff rate.
The Structural Gravity Model of Bilateral Trade Policy
While the ERP formula evaluates supply-chain price distortion, the structural gravity framework (Anderson and van Wincoop, 2003; Yotov et al., 2016) provides the empirical standard for estimating the ex-post trade volume effects of policy agreements and tariff changes. Unlike standard linear regression analysis, modern gravity modeling specifies a multiplicative exponential conditional mean:
Xij,t = exp(πi,t + χj,t + μij + γ τij,t) × εij,t
Where:
- Xij,t: Bilateral export flow from origin country i to destination country j in period t.
- πi,t: Exporter-time fixed effects, which fully capture exporter domestic production capacity and outward multilateral resistance (the average trade barrier faced by exporter i across all world markets).
- χj,t: Importer-time fixed effects, capturing importer total expenditure and inward multilateral resistance (the average price level in market j).
- μij: Country-pair fixed effects, which absorb all bilateral time-invariant geographical, cultural, and historical trade frictions (bilateral physical distance, shared borders, linguistic commonality, and colonial ties).
- τij,t: A vector of time-varying trade policy variables, including bilateral tariff changes ln(1 + Tariffij,t) and binary indicators for Regional Trade Agreements (RTAij,t).
- γ: The estimated policy elasticity parameter. For ad valorem tariffs, γ = 1 − σ, where σ represents the elasticity of substitution across product varieties.
- εij,t: A multiplicative error term estimated using Poisson regression models (specifically Poisson Pseudo-Maximum Likelihood, PPML) to accommodate zero trade flows and heteroskedastic error structures without selection bias.
Worked Example: Calculating Effective Rate of Protection with Numerical Arithmetic
To demonstrate the calculation and economic mechanics of trade policy protection, consider an economy that manufactures commercial power transformers (final good j). All monetary values and cost shares are evaluated in constant U.S. dollars at world free-trade prices.
Free-Trade Baseline Parameters
In the absence of tariffs, the world market price of a commercial transformer is Pj = $1,000. Producing one transformer requires two primary imported intermediate inputs:
- Input 1 (Specialized electrical steel sheet): Free-trade cost = $400. Input share a1j = $400 / $1,000 = 0.40.
- Input 2 (Insulated copper windings and control modules): Free-trade cost = $200. Input share a2j = $200 / $1,000 = 0.20.
- Total intermediate input cost: $400 + $200 = $600. Total input share ∑i aij = 0.40 + 0.20 = 0.60.
- Domestic value added under free trade (Vj): $1,000 − $600 = $400. Per dollar of gross output, Vj = 1 − 0.60 = 0.40 (or 40.0% of final output value).
Scenario A: Tariff Escalation Policy
The government establishes a protective trade policy tariff schedule:
- Nominal tariff on imported finished transformers: tj = 15.0% (0.15).
- Nominal tariff on imported electrical steel sheet: t1 = 5.0% (0.05).
- Nominal tariff on imported copper modules: t2 = 10.0% (0.10).
Step-by-step arithmetic computation:
- Calculate the weighted average tariff on intermediate inputs:
∑i aij ti = (a1j × t1) + (a2j × t2)
∑i aij ti = (0.40 × 0.05) + (0.20 × 0.10) = 0.020 + 0.020 = 0.040 (or 4.0% of the world final output price). - Calculate domestic prices under the tariff structure:
Domestic price of finished transformer: P′j = $1,000 × (1 + 0.15) = $1,150.00.
Domestic cost of Input 1: $400 × (1 + 0.05) = $420.00.
Domestic cost of Input 2: $200 × (1 + 0.10) = $220.00.
Total domestic input cost: $420.00 + $220.00 = $640.00. - Calculate domestic value added with tariffs (V′j):
V′j = P′j − Total domestic input cost = $1,150.00 − $640.00 = $510.00. - Calculate the change in domestic value added:
ΔVj = V′j − Vj = $510.00 − $400.00 = $110.00. - Calculate the Effective Rate of Protection (ERPj):
ERPj = ΔVj / Vj = $110.00 / $400.00 = 0.2750 (27.50%). - Verify using the direct formula:
ERPj = (tj − ∑i aij ti) / (1 − ∑i aij) = (0.150 − 0.040) / (1 − 0.600) = 0.110 / 0.400 = 0.2750 (27.50%).
Result analysis: While the nominal tariff on finished machinery is 15.0%, the domestic manufacturing activity receives an effective protection rate of 27.50% on its value added. Tariff escalation heavily magnifies the protective barrier enjoyed by the local processing plant.
Scenario B: Inverted Tariff Structure (Negative Effective Protection)
Suppose the policymaking authority alters the tariff schedule to protect domestic metal producers while seeking to lower capital costs for industrial equipment buyers: tj is lowered to 5.0% (0.05), while input tariffs are raised to t1 = 15.0% (0.15) and t2 = 10.0% (0.10).
- Weighted intermediate tariff: (0.40 × 0.15) + (0.20 × 0.10) = 0.060 + 0.020 = 0.080 (8.0%).
- Domestic price with tariff: P′j = $1,000 × (1 + 0.05) = $1,050.00.
- Domestic input costs: Input 1 = $400 × 1.15 = $460.00; Input 2 = $200 × 1.10 = $220.00. Total input cost = $680.00.
- Domestic value added (V′j): $1,050.00 − $680.00 = $370.00.
- Effective Rate of Protection:
ERPj = ($370.00 − $400.00) / $400.00 = −$30.00 / $400.00 = −0.0750 (−7.50%).
Formula check: (0.050 − 0.080) / (1 − 0.600) = −0.030 / 0.400 = −0.0750 (−7.50%).
Result analysis: Domestic transformer manufacturers face an effective tax of 7.50% on their processing operations compared to foreign competitors who purchase untaxed inputs and export finished units at the 5.0% tariff rate.
Statistical Interpretation and Decision Matrix
Interpreting trade policy indicators requires comparing nominal tariff rates against effective value-added rates, evaluating tariff dispersion, and examining binding overhang. The decision matrix below summarizes the statistical interpretation and economic implications of different Effective Rate of Protection outcomes.
| ERP Condition | Mathematical Relationship | Economic Meaning | Policy Diagnostic |
|---|---|---|---|
| Tariff Escalation | ERPj > tj (tj > ∑ aij ti) | Effective protection on value added exceeds the statutory nominal tariff. | Protects domestic downstream processing; encourages domestic manufacturing of finished goods at consumer expense. |
| Uniform Protection | ERPj = tj (tj = ti for all i) | Effective protection equals nominal protection; no relative distortion across input stages. | Neutral across processing stages; avoids unintended supply-chain penalties while generating fiscal revenue. |
| De-escalation | 0 < ERPj < tj (0 < ∑ aij ti < tj) | Positive net protection on value added, but lower than the nominal tariff on the finished good. | Intermediate input tariffs reduce the net protective margin of final goods manufacturers. |
| Negative Protection | −100% < ERPj < 0 (∑ aij ti > tj) | Domestic value added is reduced below free-trade levels by heavy intermediate input taxation. | Inverted tariff structure; penalizes domestic assemblers and induces offshoring of domestic manufacturing. |
| Extreme Inversion | ERPj < −100% (V′j < 0) | Domestic value added under prevailing tariffs becomes negative at world prices. | Severe policy distortion; domestic industry operates only through prohibitive output tariffs or direct subsidies. |
The data table below compares simple applied MFN tariffs, trade-weighted tariffs, bound tariff coverage, simple average bound tariffs, and duty-free import shares across major global trading economies, highlighting structural differences in tariff commitments and binding overhang.
Key Assumptions and Common Pitfalls in Trade Policy Modeling
Conducting empirical trade policy research requires navigating several econometric and theoretical constraints. Analysts frequently encounter methodological pitfalls that introduce severe estimation bias if not properly addressed:
- Fixed Input Coefficients Assumption (Leontief Technology): The classic Corden-Balassa ERP formula assumes that input-output coefficients (aij) remain constant when relative prices change. In reality, producers substitute away from tariff-penalized imported inputs toward domestic alternatives. Assuming zero factor substitution overstates both the protective benefit of tariff escalation and the penalty of negative ERP. Researchers evaluating large tariff shocks should supplement static ERP calculations with Computable General Equilibrium (CGE) models that incorporate Constant Elasticity of Substitution (CES) production functions.
- Omission of Multilateral Resistance Terms (MRT): Naive gravity estimations that regress bilateral trade flows on bilateral tariffs and distance using ordinary least squares without directional exporter and importer fixed effects suffer from omitted variable bias. As Anderson and van Wincoop (2003) demonstrated, trade between country i and country j depends not only on their bilateral trade costs, but on their relative costs compared to all alternative global trading partners. Omitting exporter-time and importer-time fixed effects violates structural gravity theory and produces biased policy elasticities.
- Jensen’s Inequality and Log-Linearization Bias in OLS: Historically, empirical researchers estimated gravity models by taking logarithms of trade flows: ln(Xij) = β0 + β1 ln(GDPi) + β2 ln(GDPj) − γ ln(τij) + eij. As Santos Silva and Tenreyro (2006) proved, because the expected value of the log of a random variable does not equal the log of its expected value (E[ln y] ≠ ln E[y]) under heteroskedasticity, log-linear OLS estimates are fundamentally biased and inconsistent. Furthermore, log-linearization forces researchers to drop zero-trade observations (ln(0) is undefined), discarding critical economic information about non-trading country pairs. Estimating the multiplicative gravity model using Poisson Pseudo-Maximum Likelihood (PPML) resolves both issues simultaneously.
- Endogeneity and Reverse Causality in Trade Policy Agreements: Governments do not negotiate Free Trade Agreements (FTAs) or adjust tariff schedules randomly; pairs of countries with historically strong trade connections or complementary economic structures are significantly more likely to sign preferential agreements. Failing to control for time-invariant unobserved bilateral heterogeneity using country-pair fixed effects (μij) biases policy coefficients upward. As documented across leading economics working paper series, incorporating symmetric or directional pair fixed effects is the required econometric standard to identify policy effects purely from within-pair temporal variation.
- Weighting Bias in Aggregate Tariff Indicators: When summarizing national trade policy stance, simple mean tariffs weight each tariff line equally regardless of trade volume, giving obscure product lines identical weight to major agricultural or industrial commodities. Conversely, trade-weighted tariffs weight each tariff by its observed import volume; this introduces severe downward bias because highly protective or prohibitive tariffs choke off import volumes entirely, resulting in near-zero statistical weights for the most restrictive barriers. Researchers should report both metrics alongside the percentage of duty-free tariff lines.
How to Report Trade Policy Findings in APA Format
When reporting empirical trade policy evaluations and gravity regression models in academic manuscripts, researchers should follow APA 7th edition conventions for statistical reporting, detailing model specifications, estimation techniques, clustering dimensions, and exact effect sizes.
Sample Narrative Reporting:
A Poisson pseudo-maximum likelihood (PPML) structural gravity regression was estimated to evaluate the bilateral trade effects of preferential trade agreements across 64 economies over the 2010–2024 period. The model incorporated directional exporter-by-year, importer-by-year, and country-pair fixed effects, with standard errors clustered at the pair level (N = 81,920 country-pair-year observations). The estimated coefficient for regional trade agreement membership was statistically significant, γ = 0.382, SE = 0.041, z = 9.32, p < .001, indicating an average trade creation effect of 46.5% [calculated as (exp(0.382) − 1) × 100] between signatory partners, holding multilateral resistance and bilateral gravity determinants constant. The bilateral applied tariff elasticity was γtariff = −4.180, SE = 0.450, z = −9.29, p < .001, implying that a 1% reduction in bilateral ad valorem tariffs increased bilateral merchandise exports by approximately 4.18%.
| Model Variable | Coefficient (γ) | Robust Clustered SE | z-Statistic | p-Value | 95% Confidence Interval |
|---|---|---|---|---|---|
| Regional Trade Agreement (RTA) | 0.382 | 0.041 | 9.32 | < .001 | [0.302, 0.462] |
| Log Applied Tariff [ln(1 + Tariff)] | −4.180 | 0.450 | −9.29 | < .001 | [−5.062, −3.298] |
| Exporter × Year Fixed Effects | Yes | — | — | — | Controlled |
| Importer × Year Fixed Effects | Yes | — | — | — | Controlled |
| Country-Pair Fixed Effects | Yes | — | — | — | Controlled |
How to Calculate Trade Policy Metrics in Excel, R, and Python
Below are verified, executable computational routines for calculating the Effective Rate of Protection (ERP) and setting up empirical trade policy estimations in spreadsheet software and statistical programming environments.
Calculating ERP in Microsoft Excel
To calculate Effective Rate of Protection in Excel for a product with multiple intermediate inputs, structure your worksheet as follows:
- Cell
B1: Nominal tariff on final output (e.g.,0.15for 15%). - Cells
A4:A5: Input names (e.g., Steel Sheet, Copper Modules). - Cells
B4:B5: Free-trade cost shares (e.g.,0.40inB4,0.20inB5). - Cells
C4:C5: Input nominal tariffs (e.g.,0.05inC4,0.10inC5). - Cell
B6(Total input share):=SUM(B4:B5) - Cell
B7(Weighted intermediate tariff):=SUMPRODUCT(B4:B5, C4:C5) - Cell
B8(Free-trade value added):=1-B6 - Cell
B9(Effective Rate of Protection):=(B1-B7)/B8
Evaluating this spreadsheet structure with the worked example parameters yields exactly 0.2750 (27.50%).
Calculating Trade Policy Metrics in R
In R, researchers calculate the Effective Rate of Protection using vectorized functions and estimate structural gravity models using the high-performance fixest package:
# Function to calculate Effective Rate of Protection (ERP)
calculate_erp <- function(t_final, a_shares, t_inputs) {
if (length(a_shares) != length(t_inputs)) {
stop("Input cost shares and input tariffs must have equal length.")
}
total_share = 1.0) {
stop("Sum of input cost shares must be strictly less than 1.0.")
}
weighted_input_tariff <- sum(a_shares * t_inputs)
free_trade_va <- 1.0 - total_share
erp <- (t_final - weighted_input_tariff) / free_trade_va
return(erp)
}
# Example execution using worked example parameters
erp_val <- calculate_erp(
t_final = 0.15,
a_shares = c(0.40, 0.20),
t_inputs = c(0.05, 0.10)
)
cat(sprintf("Effective Rate of Protection: %.4f (%.2f%%)n", erp_val, erp_val * 100))
# Structural Gravity PPML Estimation syntax using fixest
# library(fixest)
# gravity_model <- fepois(
# exports ~ log_tariff + rta | exp_id^year + imp_id^year + pair_id,
# data = trade_panel_df,
# cluster = ~ pair_id
# )
# summary(gravity_model)
Calculating Trade Policy Metrics in Python
The standard Python implementation provides clean, dependency-free calculation of Effective Rate of Protection using built-in iterables, suitable for integration into data processing pipelines:
def calculate_effective_rate_of_protection(
t_final: float,
input_shares: list[float],
input_tariffs: list[float]
) -> float:
# Validate identical lengths
if len(input_shares) != len(input_tariffs):
raise ValueError("input_shares and input_tariffs must have identical length.")
total_share = sum(input_shares)
if total_share >= 1.0:
raise ValueError("Sum of input cost shares must be strictly less than 1.0.")
weighted_input_tariff = sum(s * t for s, t in zip(input_shares, input_tariffs))
free_trade_value_added = 1.0 - total_share
erp = (t_final - weighted_input_tariff) / free_trade_value_added
return erp
# Execute with worked example parameters
erp_result = calculate_effective_rate_of_protection(
t_final=0.15,
input_shares=[0.40, 0.20],
input_tariffs=[0.05, 0.10]
)
print(f"Effective Rate of Protection: {erp_result:.4f} ({erp_result * 100:.2f}%)")
# Output: Effective Rate of Protection: 0.2750 (27.50%)
Frequently asked questions
What is the difference between nominal tariffs and the effective rate of protection?
A nominal tariff is the statutory customs duty rate levied on the total gross invoice value of an imported product at the border. In contrast, the Effective Rate of Protection (ERP) measures the net percentage increase in domestic value added per unit permitted by the full tariff schedule, explicitly accounting for customs duties imposed on imported intermediate inputs, components, and raw materials used in production.
What is tariff escalation and why is it important in trade policy?
Tariff escalation occurs when an importing economy levies lower tariffs on unprocessed raw materials, moderate tariffs on intermediate goods, and higher tariffs on finished manufactured goods. Because the final good tariff exceeds input tariffs, the effective rate of protection on domestic processing value added significantly exceeds the nominal tariff rate, creating strong economic incentives for domestic manufacturing while discouraging value-addition in exporting trading partners.
How does binding overhang affect trade policy predictability?
Binding overhang represents the positive gap between a WTO member’s legally bound tariff ceiling committed under GATT Article II and its lower applied Most-Favoured-Nation (MFN) tariff. When a large binding overhang exists, governments retain legal latitude under international agreements to abruptly increase applied duties up to the bound rate without renegotiation, generating substantial trade policy uncertainty that deters long-term cross-border business investment.
Why is Poisson Pseudo-Maximum Likelihood (PPML) preferred over OLS for trade policy estimation?
In empirical trade policy modeling, Poisson Pseudo-Maximum Likelihood (PPML) estimation is preferred over log-linear ordinary least squares because it overcomes Jensen’s inequality bias in the presence of heteroskedastic error terms and naturally incorporates zero bilateral trade observations without artificial data adjustments. Furthermore, PPML ensures that the estimated fixed effects preserve the structural multilateral resistance properties required by modern gravity theory.
How do non-tariff measures (NTMs) differ from tariffs in quantitative analysis?
Unlike customs tariffs, which are transparent monetary taxes collected at the border, non-tariff measures (NTMs) consist of regulatory requirements, conformity assessment procedures, sanitary and phytosanitary rules, and technical standards. In quantitative analysis, NTMs cannot be measured as statutory tax percentages; economists must calculate import coverage ratios, frequency indices, or estimate ad valorem equivalents (AVEs) using price-gap analysis or econometric gravity equations.
Sources
- World Trade Organization and UNCTAD. (2012). A Practical Guide to Trade Policy Analysis. World Trade Organization. https://www.wto.org/english/res_e/publications_e/practical_guide12_e.htm (checked 2026-11-14).
- World Trade Organization and UNCTAD. (2016). An Advanced Guide to Trade Policy Analysis: The Structural Gravity Model (by Yotov, Y. V., Piermartini, R., Monteiro, J.-A., and Larch, M.). World Trade Organization. https://www.wto.org/english/res_e/publications_e/advanced_guide_tr_policy_art16_e.htm (checked 2026-11-14).
- World Trade Organization. (2026). WTO: Tariffs. World Trade Organization. https://www.wto.org/english/tratop_e/tariffs_e/tariffs_e.htm (checked 2026-11-14).
- World Trade Organization, UNCTAD, and International Trade Centre. (2026). World Trade Organization: International Trade and Tariff Data. World Trade Organization. https://www.wto.org/english/res_e/statis_e/statis_e.htm (checked 2026-11-14).
- World Bank, UNCTAD, WTO, and UNSD. (2026). World Integrated Trade Solution (WITS) Database. World Bank. https://wits.worldbank.org/ (checked 2026-11-14).
- World Trade Organization. (2026). World Trade Report. World Trade Organization. https://www.wto.org/english/res_e/publications_e/wtr_e.htm (checked 2026-11-14).
Data: Global Trade Policy and Tariff Profile Comparison
Cross-country comparison of trade policy metrics, including simple average MFN applied tariffs, trade-weighted applied tariffs, bound tariff coverage, simple average bound tariffs, and duty-free import shares, compiled from WTO World Tariff Profiles and World Bank WITS data.
Period covered: 2024/2026. Coverage: Global.
| Economy | Simple Average Applied Tariff (%) | Trade-Weighted Applied Tariff (%) | Bound Tariff Coverage (%) | Simple Average Bound Tariff (%) | Duty-Free Imports Share (%) |
|---|---|---|---|---|---|
| United States | 3.4 | 2.3 | 100 | 3.4 | 69.2 |
| European Union | 5.1 | 2.8 | 100 | 5.1 | 64.8 |
| Japan | 4.2 | 2.2 | 99.7 | 4.5 | 72.1 |
| Canada | 3.9 | 1.6 | 99.7 | 6.5 | 78.4 |
| Australia | 2.4 | 1.2 | 97.2 | 9.8 | 86.5 |
| United Kingdom | 3.8 | 2.1 | 100 | 5.6 | 68.4 |
| South Korea | 13.6 | 7.8 | 94.6 | 16.5 | 38.2 |
| Mexico | 6.7 | 2.1 | 100 | 36.1 | 61.5 |
| Brazil | 13.3 | 9.8 | 100 | 31.4 | 32.1 |
| India | 18.1 | 12.6 | 74.3 | 50.8 | 29.4 |
Sources
- World Trade Organization: International Trade and Tariff Data (accessed 2026-11-14)
- World Integrated Trade Solution (WITS) Database (accessed 2026-11-14)
Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.







