I am overwhelmingly against protectionism, like almost every economist. This even extends to labor, where I'm against immigration restrictions on people who can work productively here (more on this in a future post).
However, I've been thinking, and there's one kind of good that protectionism may not be so bad on. With goods where there's no difference in utility provided by the good between the American and the foreign version, and there's no difference in cost of the good (or the foreign one is more expensive), but because of heavy branding, the foreign one still sells better, then protectionism is a way of forcing perfect information, and can be a good thing (BEFORE you factor in retributive tariffs, which will always be a problem and probably mean that this is stuck in the realm of theory).
An example would be something like perfume.
1) It is highly unlikely that there are any major "quality" differences in perfumes across countries, because the quality is entirely observable in the scent, and the base is alcohol, which is universal. There shouldn't be any hidden quality variables (like health effects or how long the scent lasts) with various different perfumes that aren't the same for American and foreign varieties.
2) It is also highly unlikely that there are major cost differences in perfumes. Though I'm no perfume expert, I'd imagine perfume production costs are largely commodity raw materials and capital costs, with fairly low labor costs. Any slightly higher labor costs in the US are probably offset by the fact that we have some of the lowest capital costs in the world due to our financial-system-until-2006 and our advanced science and scientists. Commodities are commodities and cost the same roughly year round.
3) Branding is HUGE in perfume - A bottle of the original Chanel scent can cost almost $1000, while cheaper varieties made in America can cost 1/10 that. Even if Chanel's scent is distinctive, you probably can come up with a close substitute without violating patent laws. People who buy Chanel are either doing so for the "cache" and bragging rights or because they don't know such close substitutes exist.
Therefore, by tariffing Chanel perfume, you reduce consumption of Chanel (by a little bit), raise money for the government, and may actually increase the overall utility of the people who are forced to switch from Chanel to lower-quality perfume, even if they don't know it. You provide them with $800 of extra disposable income without really affecting the quality of perfume they have. Some portion of that $800 will get spent on American goods, so America ends up being better off for that tariff.
This is a highly unusual circumstance (I'd imagine it's mostly consumer goods, mostly upscale, mostly with low labor component of production), and ignores retributive tariffs, which is a game theory issue and probably still means tariffs are a bad idea. However, if you could guarantee no retributive tariffs, these criteria constitute (probably) the one type of tariff that makes people better off. I am no perfume expert, but it seems like the type of good that meets the criteria.