When trade policy changes 53 times in a year, the uncertainty can cost as much as the policy itself. That is the finding of Tariff Confusion, a discussion paper from LSE's Centre for Economic Performance by Kalina Manova, Dennis Novy, Thomas Sampson and Aaron Tang. The average statutory US import tariff rose from 4.9 per cent in January 2025 to 20 per cent in December 2025 through a cascade of executive orders and proclamations that left importers struggling to establish which rate actually applied.
Because confusion cannot be observed directly, the researchers proxy it using the number of announcements importers faced and the gap between the maximum announced tariff and the tariff actually implemented. Their estimates show that by December 2025, higher tariffs had reduced US imports by 4 to 6 per cent on average across country-product pairs compared with a year earlier, while tariff confusion had reduced imports by an additional 7 to 12 per cent.
The mechanism, the authors argue, is that exporters and importers could not update their beliefs about statutory tariffs quickly enough, and feared rates might exceed the cut-off levels at which trade remains profitable, so they simply curtailed activity. The results provide rare quantitative evidence that how trade policy is implemented, not just its level, shapes its economic consequences, a lesson with obvious relevance for any government contemplating rapid-fire protectionism.