From liberalization to fragmentation: new evidence on global tariff dynamics in the WTO era

Key findings from this study

• The study found that global tariffs converged significantly between 1990 and 2020, with initially high-tariff countries reducing rates faster than peers and cross-country dispersion declining steadily.
• The researchers demonstrate that major institutional reforms in the European Union, Japan, and China produced distinct structural breaks in liberalization trajectories aligned with integration events.
• The study found that the United States experienced tariff divergence after 2018, marking the first major reversal in global convergence and introducing fragmentation driven by geopolitical tensions rather than economic fundamentals.

Overview

This study investigates global tariff convergence between 1990 and 2020 using data from five major economies and the world average. Analysis reveals sustained long-term liberalization characterized by higher-tariff countries aligning with lower protection levels. The research identifies a significant structural break beginning in 2018, marked by U.S. tariff escalation that reverses the established convergence trajectory and introduces fragmentation into the multilateral trading system.

Methods and approach

The research employs a mixed empirical strategy combining beta-convergence and sigma-convergence tests to assess tariff alignment dynamics. Structural break analysis using Quandt-Andrews unknown breakpoint tests identifies major policy turning points. A difference-in-differences estimation quantifies asymmetric effects of the 2018 U.S. tariff escalation. Data cover the United States, China, Japan, the European Union, and world average from 1990 to 2020. Driscoll-Kraay standard errors address cross-sectional dependence and serial correlation in the panel specification.

Results

Beta-convergence tests confirm that countries with initially higher tariff levels reduced rates more rapidly, driving alignment with more liberal regimes. Sigma-convergence analysis documents steady decline in cross-country tariff dispersion, though convergence velocity slowed after the 2000s. Structural break tests identify critical liberalization episodes: European Union (1993), Japan (1995), and China (1994 and 2001), corresponding to major institutional reforms and integration events.

The United States demonstrates clear divergence after 2018, representing the first major reversal in global tariff alignment trajectories. Difference-in-differences estimates show U.S. tariffs increased significantly relative to other major economies during this period. This post-2018 rupture marks a transition from sustained multilateral liberalization to fragmented protectionism driven by geopolitical tensions rather than economic fundamentals.

Implications

The confirmed long-term convergence trend underscores the importance of international coordination and institutional frameworks—particularly the WTO—in sustaining trade openness. Structural breaks linked to major institutional reforms demonstrate that policy architecture accelerates liberalization episodes. However, the identified post-2018 divergence signals that uncoordinated protectionism disrupts policy alignment, increases supply-chain uncertainty, and threatens the coherence of the multilateral trading system. Strengthening multilateral mechanisms and improving regional trade agreement compatibility are essential to restore predictability.

Shifts in tariff policy carry distributional consequences. The sustained convergence period supported consumer welfare through lower prices and broader product variety, particularly benefiting developing economies integrating into global value chains. The recent divergence reintroduces uncertainty disproportionately affecting workers in trade-dependent sectors and export-oriented economies. Supply-chain disruptions may increase household living costs. Stable, predictable trade policies are therefore critical for social cohesion and inclusive economic opportunities.

Scope and limitations

This summary is based on the study abstract and available metadata. It does not include a full analysis of the complete paper, supplementary materials, or underlying datasets unless explicitly stated. Findings should be interpreted in the context of the original publication.