The year 2025 was unprecedented in the frequency, complexity, and volatility of US trade policy interventions. This column uses a new database tracking US tariff announcements during the year to show that confusion about the current state of tariff policy led to a decline in US imports that was similar in size to the reduction caused by tariff increases alone, meaning that confusion doubled the negative impact of tariffs. Product customisation, relationship trust, and partner reputation can help firms maintain trade activity during information shocks.
2025 witnessed a series of major changes in US trade policy. From new steel and aluminium duties to “Fentanyl” tariffs on China and “Liberation Day” tariffs on almost all trade partners, the year was unprecedented in the frequency, complexity and volatility of US policy interventions. Between February and December 2025, 53 announcements introduced, delayed, reinstated or changed tariffs, with different countries and products variably affected and intricate stacking rules and exemptions applied.
Overall, these announcements raised tariffs, and it is well-established that higher tariffs reduce trade (Alfaro and Chor 2025). But unchartered policy turbulence raises a critical question: Can frequent, complex tariff revisions generate sufficient confusion about prevailing duties to cause additional trade disruption, beyond the direct impact of tariffs?
Business reactions in 2025 certainly indicated rising confusion among firms worldwide. “Tariffs mean an economic hit for US firms – but also confusion,” augured Fortune in April. “The economy is slowing down as companies can’t figure out the rules of this tariff game,” alerted Investopedia in June. “Confused and concerned, CEOs get to grips with [the] new tariff regime [and] a new era of tariff complexity,” flagged CNBC in August, quoting the CEO of German insurance giant Allianz, “If you are not confused, I don’t know what confuses a human being”. The Cato Institute capped December with “Welcome to tariff complexity hell […] Complexity is a tax, and US companies are paying through the nose.”
In recent work (Manova et al. 2026), we propose that tariff confusion about current policy can have first-order detrimental effects on trade. We develop the first indicators of tariff confusion, document its impact on trade, and inform its operating mechanisms in the context of US trade policy in 2025.
We construct the new US Tariff Announcement Database (USTAD), which tracks US tariff announcements during 2025. We use USTAD to compute monthly US statutory tariffs and novel measures of tariff confusion by origin country and product: the number of relevant announcements, the number of possible tariff calculations arising, and bounds on possible tariff miscalculation.
We find that confusion about the current state of US tariff policy led to significant declines in US imports during 2025. Strikingly, we estimate that the fall in US imports due to tariff confusion was similar in size to the reduction caused by tariff increases alone, meaning that confusion doubled the negative impact of tariffs. Notably, tariff confusion was persistent and more damaging at higher tariff levels, but less disruptive for relationship-specific goods and origins with stronger trust systems. These results point to novel consequences of the manner in which trade policy changes are implemented.
We conceptualise tariff confusion as confusion about the contemporaneous statutory tariff on imports of a specific product from a given country. Our premise is that periods of frequent and complex policy announcements require firms to continuously monitor and process the arrival of new information, and that this information problem becomes more difficult as announcements accrue over time. Tariff confusion may thus arise because foreign exporters and/or US importers miss an announcement, do not recognise its relevance to their origin-product, or struggle to determine how it affects the tariff rate. While distinct from uncertainty about hypothetical future tariffs (Handley and Limao 2022), confusion about actual current tariffs can be seen as a manifestation of trade policy uncertainty in real time.
US trade policy in 2025 featured arguably the first episode of widespread tariff confusion in a modern economy. Two examples illustrate its complexity and dynamism. First, in June, cars were newly exempt from the acting 10% global reciprocal tariff. EU cars thus incurred a new blanket 25% automobile tariff. However, Canadian USMCA-compliant cars faced no duty, while non-compliant cars received a flat 25% tariff on Canada. And Chinese cars bore both the 25% car tariff and the 20% China tariff, for a total of 45%. Second, between 5 April and 14 May, the duty on Chinese shoes over the most-favoured nation (MFN) rate moved from 20% (“fentanyl”) to 30% (+ 10% global reciprocal tariff), 54% (+ 34% China tariff), 104% (+ 84% China retaliatory tariff), 145% (+ 125% second China retaliatory tariff), and back to 30% (+ 10% global reciprocal tariff).
Consistent with tariff confusion, a survey of 4,400 US and Canadian firms in March-April 2025 revealed that many were imperfectly informed (Atkin et al. 2025). For example, 45% of firms believed that tariffs on Chinese imports were below 20%, even after announcements had set the minimum to 20% above MFN, or 42% on average across products.
We record all 53 US trade policy announcements made through presidential executive orders and proclamations in 2025. These announcements led to 1.45 million changes to tariffs on origin-products the US imported in 2025, including 1.05 million rises and 0.39 million cuts.
Using USTAD, we calculate the statutory import Tariff and four indicators of potential tariff confusion by country-product-month. We measure the stock of information that required monitoring with the cumulative number of relevant announcements that affect a given origin-product (#Announce). As for processing policy updates, we proxy the scope for inaccurate tariff calculations with the number of possible announcement combinations firms may track (TariffMess), the highest possible tariff they may compute (TariffMax), and the gap between this highest tariff and actual tariffs (TariffMiss).
The statutory tariff and our tariff confusion measures all increased sharply during 2025, with considerable variation across origin country-product pairs (Figure 1). Moreover, tariff confusion is not strongly correlated with statutory tariffs: in December, Tariff’s correlation with #Announce, TariffMess and TariffMiss was as low as 0.24, 0.04, and 0.24, respectively.
Figure 1 US statutory tariffs and tariff confusion grew sizably and variably across countries and products over January – December 2025
We exploit the variation in statutory tariff changes and tariff confusion across products, origin countries and months during 2024-25 to estimate the short-term impact of US policy changes on US imports. We account for foreign supply potential and US supply/demand conditions with appropriate fixed effects.
Both higher US tariffs and greater tariff confusion significantly disrupted US imports in 2025. A 10% higher tariff reduced monthly imports by 3.3-4.9% on average across countries and products. Conditional on the tariff, each additional tariff announcement lowered imports by another 3.0%.
Our estimates imply that higher US tariffs reduced US imports in December 2025 relative to December 2024 by 4-6% for the average origin country-product pair. Cumulative tariff confusion cost an additional 7-12% import fall. This additional fall would not have occurred had tariffs been raised once and for all in a clear policy statement.
The incidence of US trade policy varied across countries (Figure 2). We estimate that the average country exported 13% less to the US in the month when it faced the highest tariffs, of which 7% due to confusion arising from multiple announcements. The estimated fall in exports ranges dramatically from 0.4% to 57% across countries, but tariff confusion always contributes substantially to the decline.
Figure 2 US import decline at peak country tariff varied across countries, with tariff confusion consistently contributing significantly
The consequences of tariff confusion were neither transitory nor dampening over time, with both recent and three-month lagged confusion exerting significant, negative effects on US imports. We also find that confusion was more damaging at higher tariff rates, and that its impact cannot be attributed to uncertainty about future trade policy measured using Liberation Day tariffs as a threat point.
Results also indicate that US importers bore the burden of tariff increases and tariff confusion. The reductions in US imports caused by higher tariffs and tariff confusion both occurred primarily through lower import quantities, with little or no effect on pre-tariff import prices.
Notably, US imports of products that are differentiated, require more relationship-specific investments, or have stickier buyer–supplier relationships were more resilient to tariff confusion (Figure 3). US imports from countries with greater trust in foreigners were also less affected. These findings are consistent with product customisation, relationship trust, and partner reputation enabling firms to maintain trade activity during information shocks.
Figure 3 Trade was more resilient to tariff confusion for relationship-specific goods and origins that trust foreigners more
Our results indicate that policy confusion has first-order effects on firms’ decision making. A single, unambiguous policy statement is less disruptive than repeated, complex announcements. The drop in US imports caused by the 2025 tariffs would have been around 50% smaller if there had been no confusion about US tariff policy. Better understanding how firms update beliefs and manage buyer-supplier relationships to adapt to policy confusion would be a fruitful agenda for future research. Policy confusion may also exert different medium-to-long run effects depending on the duration and ultimate resolution of policy turbulence, as well as on firms’ prior experience with it.
Finally, our findings highlight the importance of the design of the trade policy process. Monetary authorities have long recognised the value of transparency and predictability, and perhaps these objectives should guide national and multilateral policy making more generally.
Source : VOXeu
US net migration turned negative in 2025 following a sharp tightening of immigration policies. This…
Central banks can move policy rates, but transmission depends on households acting on them. This…
Rising bond yields are likely down to stronger investment demand, more supply of long-duration debt…
Falling agricultural prices have been found to generally increase armed civil conflict in producing regions,…
The White House Faith Office is expanding federal support for religious organisations. Can government policy…
Geopolitical tensions have reshaped international trade and investment flows in recent years. This column uses…