A tariff is a tax on imports, paid by the importer at the port of entry. That part is mechanical. The harder question, and the one that decides whether a tariff shows up in CPI, is who eats the cost.
There are three places it can land. The exporter can cut its price to keep the sale, absorbing the tariff abroad. The importer or retailer can absorb it in their margin. Or the price can pass through to the consumer. In practice, most large tariffs split across all three, and the split depends on competition.
For commodities with many global suppliers, exporters tend to absorb little. The retailer cannot squeeze the producer because the producer can sell elsewhere. Pass-through to consumers is high, often 70 to 90 percent within a few months.
For branded goods where the importer has pricing power and the consumer has alternatives, the retailer absorbs more. Pass-through is lower and slower, often 30 to 50 percent over a year. The lag matters: retailers work through inventory bought at the old price first, which is why tariff effects show up in CPI months after the tariffs take effect.
There is also a substitution channel. When tariffs raise the price of imported washing machines, domestic washing machine prices rise too, because the import competition that was disciplining domestic pricing is gone. Economists call this the indirect effect. It is often larger than the direct effect on imported goods.
Tariffs are not a one-time price level shock unless they are clearly temporary. If markets expect them to persist, supply chains reorganize, domestic capacity gets built, and the new price level becomes the baseline. That is what makes the inflation question hard: tariffs change the level once, but the level shift can be substantial, and the second-round effects on wages and other prices can extend the impact well beyond the first year.
The cleanest read on tariff pass-through is import prices ex-tariffs, published by BLS each month. When that index rises, exporters are not absorbing. When it falls, they are.