S9 Ep45: Tariff Confusion

Episode Summary

Executive Summary: The episode examines a 2025 U.S. tariff regime defined by constant changes, complex stacking rules, and unclear implementation. Kalina Manova argues that confusion about tariffs acted like an additional tax on trade: firms often misread rates, and the resulting uncertainty reduced U.S. imports about as much as, or more than, the tariffs themselves.

Main Topics: Unprecedented tariff volatility in 2025 (Priority: 5/5): The U.S. changed tariff policy through frequent executive orders and proclamations affecting countries and products, creating an unusually unstable policy environment compared with the last 150 years. Why tariff confusion arose (Priority: 5/5): Tariffs were introduced, delayed, reinstated, and stacked in ways that made it difficult for firms to know the true duty on a shipment, especially when rules varied by country of origin and product classification. Measuring confusion and tariff levels (Priority: 5/5): The authors built a new database of tariff announcements and created proxies for confusion based on announcement counts, possible tariff calculations, worst-case tariff exposure, and distance from actual statutory rates. Confusion reduced trade as much as tariff hikes (Priority: 5/5): Empirical results show that tariff confusion significantly lowered imports and often doubled the effect of tariff increases, demonstrating that informational friction can be as damaging as the tax itself. Heterogeneous effects across countries and products (Priority: 4/5): Trade disruptions were smaller where buyer-supplier relationships were stickier, where trust in foreigners was higher, and where product-specific contracting was more important. Persistence and policy lessons (Priority: 4/5): Confusion did not fade quickly; it persisted over months. The episode suggests that if governments want tariffs to generate revenue, policy must be stable and clearly communicated.

Key Arguments: Tariff confusion itself functions like a tax on trade, but unlike a tariff it raises no revenue. The 2025 U.S. tariff episode was unprecedented in frequency, complexity, and legal/political form over the last 150 years. Firms were genuinely confused, as shown by media reports and survey evidence on misperceptions of actual tariff rates and the number of policy changes. A new database was necessary because standard tariff reporting systems did not cleanly reflect the new announcements and often dumped them into a catch-all category. Confusion and tariff levels can be separated empirically because the two measures are only weakly correlated in many parts of the sample. The main result is that tariff confusion roughly doubled the negative trade effect of tariff hikes, and by year-end confusion reduced imports substantially beyond the tariff effect itself. The impact was smaller in markets with stronger relational contracting, more product-specific investment, and higher trust in foreigners. Confusion effects were persistent over time and not just a short-lived adjustment shock. A stable, single announcement would likely maximize tariff revenue compared with repeated revisions that generate uncertainty and lower trade further.

Data Points: Number of U.S. tariff announcements: 53 - Announcements between February and December 2025 introducing, delaying, reinstating, or changing tariffs. Period covered by announcements: 11 months - The 53 tariff announcements occurred over February to December 2025. Product-level revisions: 32 - Revisions affecting products within the 11-month period. Comparison with normal periods: 3x regular period; 2x China-US trade war - Product-level revisions were about three times a regular period and twice the 2018-2019 China-US trade war. Survey sample size: About 4,500 firms - US and Canada firms surveyed in March-April 2025 about tariff perceptions. Firms underestimating China tariff: 45% - Average share of firms that thought the tariff on goods from China was under 20% when it was about 42%. Actual China tariff level referenced: About 42% - The podcast cites this as the actual level firms failed to recognize in the survey discussion. Firms underestimating liberalizing announcements: 87% - Most firms underestimated the number of announcements that postponed or lowered tariffs. Data coverage of tariff database: 230 origin countries and over 18,000 10-digit product categories - The authors’ U.S. Tariff Announcement Database for 2025. Tariff announcements impact on imports: A 10% tariff hike lowered imports by 3-5% - Average monthly effect estimate for tariff increases. Confusion effect per additional announcement: Another 3% lower imports - Holding the tariff level constant, one more tariff announcement reduced imports by about 3%. Year-end tariff effect on imports: 4-6% lower - US imports in December 2025 versus December 2024 due to higher tariffs. Year-end confusion effect: 7-12 percentage points additional reduction - Confusion produced an extra decline beyond the tariff effect by December 2025. Peak-tariff country-level effect: 13% lower exports on average - At each country’s highest peak U.S. tariff month, exports to the U.S. were 13% lower on average. Share of that peak effect due to confusion: About half - Confusion accounted for roughly half of the 13% reduction at peak tariff exposure. Correlation between tariff max and statutory tariff early in period: About 0.5 by December - Correlation fell as exemptions and reversals accumulated. Correlation between announcement count and tariff mess: About 0.14-0.15 overall - Weak correlation between cumulative announcements and potential tariff calculations in the full panel. Cross-sectional correlation by December: Under 0.1 - Announcement count and tariff mess were very weakly correlated by year-end.

Pivotal Quotes: "If you are not confused, I don't know what confuses a human being." — Head of Allianz: Cited by Kalina Manova as an illustrative media example of business confusion over tariff rules. "Tariff confusion, on average, doubled the detrimental effect of the tariff hikes themselves." — Tim Phillips summarizing Manova's findings: The host restates the paper’s central empirical result. "Make up your mind and stick to it." — Kalina Manova: Policy lesson that stable, clear tariff announcements would reduce confusion and improve revenue collection.

Implications: The episode suggests that unstable trade policy can depress trade as much as tariff levels do. Firms need clear, durable rules; otherwise uncertainty becomes an extra hidden tax. Governments seeking revenue should prioritize stability over repeated revisions.