When Tariff Averages Misrank Markets: A Business Intelligence

Audit of WTO 2025 Profiles

Seyfettin Artan1,*

1 Department of Economics, Karadeniz Technical University, Trabzon, Turkiye

Received: August 02, 2026 Revised: September 14, 2026 Accepted: October 07, 2026 ⋆ Corresponding author

ABSTRACT

Procurement teams often rank destination markets using a single published tariff average. Whether that ordering

survives a change of measurement is a practical question for business intelligence. We audit the 2025 World

Tariff Profiles, retaining 101 economies with matching 2024 most-favoured-nation (MFN) applied tariffs, reported

trade-weighted tariffs, and 2023 import totals. A reproducible screen compares simple and trade-weighted averages,

ranking changes, duty-free subheading shares, and the share of subheadings facing duties above 15%. The median

simple average is 7.5% and the median trade-weighted average is 6.9%; the within-economy weighted-minus-simple

median is −0.6 percentage points. Thirty-five economies differ by at least two points. With ties retained at each

top-25 cutoff, 28 economies qualify under the simple rate and 26 under the weighted rate, but only 20 qualify under

both. Türkiye falls from 17.3% to 7.0%, whereas Dominica rises from 10.7% to 15.9%. Disagreement persists

among larger import markets: 13 of 31 economies with at least US$50 billion in 2023 imports have a gap of two

points or more. A layered dashboard should therefore trigger product-level verification where aggregate measures

disagree. Neither average identifies the duty paid by a particular firm or the effect of changing trade policy.

Keywords: Business intelligence Tariff analytics Market screening Import sourcing Decision support

1. INTRODUCTION

A firm selecting a sourcing location or export market may

begin with a table of national tariffs. The table compresses

thousands of product lines into one number per destination,

saving the analyst’s time at the first screening stage. But a

ranking built on a simple average of tariff lines can differ

from a ranking that gives more weight to products a country

actually imports. Neither statistic is a transaction quote.

When procurement analysts mistake either for a landed-cost

forecast, a seemingly inexpensive market may be assigned

too little attention, or a viable market may be excluded too

early.

The decision question is specific: how often does a firm’s initial

market-priority list change when a simple tariff average

is replaced by the published trade-weighted average? We

audit that choice in a defined set of economies. The resulting

diagnostic identifies where analysts should request a productlevel

rate; it cannot estimate the rate payable on a particular

shipment.

2. RELATED WORK AND RESEARCH GAP

Business analytics research emphasizes the path from integrated

data to a defensible decision. In a survey of 305 firms,

Szukits and Móricz [1] associate perceived data quality and

managerial support for analytics with an analytical decision

culture; Liu and Yoon [2] design a goal-driven method for

integrating data around a user’s decision need. These contributions

motivate a transparent sourcing dashboard, but neither