No, that's not what is happening. From the paper:

>The reason is that markups along the chain of intermediation between importer and consumer can scale up the percent pass-through in tariff costs, cumulating over distribution stages and resulting in a direct dollar impact on prices to be greater than tariffs paid, even though the percent change in consumer price is less than the tariff ad-valorem rate.

The retailer is paying more for its stock. Everyone loses despite higher prices.

Ah I see, I had misunderstood, the discrepancy in tariff revenue and what the consumer pays is due to the markups at each stage of distribution successively absorbing minor portions of the price increase until the consumer bears most of it. Took me a while to get my head around the numbers, but makes sense now, thanks for the correction!