The administration announced new tariffs on April 2. Prior to this we estimate that the effective tariff rate—tariffs collected divided by goods imports—had already increased from 2.3% in 2024 to 10%. That is on a static basis, which assumes no changes in import composition. With the latest announcement that rate would climb to 23% when fully implemented on April 9, equalling rates last seen in the early 20th century (Figure 1), and this is likely to increase further to 25% assuming that tariffs are placed on some additional products.

The latest action, which was legally justified using IEEPA, raises tariff rates by at least 10% on all countries except Canada and Mexico, with close to 60 countries seeing higher rates, including 20% or more for the largest importers (Table 1). This does not apply to steel, aluminum and autos, where separate 25% tariffs were already being implemented, or to copper, pharmaceuticals, semiconductors, lumber articles, certain critical minerals, and energy, many of which could also get a 25% tariff soon. While billed as “reciprocal tariffs” no attempt was made to measure foreign trade barriers, and instead the incremental increase was based solely on bilateral trade deficits. The formula used is:

Where exports and imports refer to 2024 goods trade, E = -4 is the elasticity of import demand to import prices, and P=0.25 is the pass-through of tariffs to import prices (see USTR). This reduces to (1-exports/imports)/2, or half the percent reduction in imports needed to close the goods trade deficit for each individual country, with a separate 10% floor. The USTR notes low pass-through from import prices to retail prices using Cavallo et al, though that same paper finds the pass-through from tariffs to import prices (P) is near 1. It’s not clear if the USTR had meant to set P = 1 (which would reduce all the rate increases by a factor of 4), though admittedly their demand elasticity assumption (E) is high, and a short-run value of -1 could be justified, which would leave the result unchanged even if P were changed to 1. In addition, the scaling by 0.5 is not mentioned in the USTR discussion.

Table 2 shows the walk from the effective tariff rate in 2024(2.3%) to the new tariff rate following the April 2 announcement (23.3%), as well as the impact of adding 25% tariffs to some of the other products that President Trump has mentioned. As a simplification we assume there is no change in the import mix, whereas in reality there will likely be attempts to substitute toward lower-tariff imports, which would reduce the increase.

These tariffs are likely to some degree a negotiating tool to bring down foreign trade barriers and we could see deals announced in coming months that reverse some of the tariffs. However, their structure also highlights the risk that tariff rates will stay high for long. First, no country outside North America was spared at least a 10% increase, suggesting this could be new global floor. Second, the rates are calibrated to bilateral trade deficits, long a focus of President Trump, which could be hard to change quickly, not to mention bilateral deficits can exist even in the absence of any barriers. Third, while import substitution, carve-outs, and payments to firms hurt by retaliation could reduce tariff revenue, the static cost of 20% increase in tariffs is more than $600bn per year, which could become an important mechanism to try to help pay for other tax cuts now being debated in Congress.