If you buy the foreign t-shirt there's $5 more floating around in the economy for you to spend. Plus you have a t-shirt. Plus the local country is likely to stop making t-shirts (which it manifestly is not competitive at) and put that money into something that it is competitive at i.e. away from something that is not productive and into something that is productive which is to say efficient which is better for everyone - even better for the environment as resources are not wasted by inefficiently making goods.
The onus is on the other side which claims that free trade is always beneficial, and any tariffs or protectionism is harmful.
Their analysis seems to work with the assumption that every participant in the economy practices free trade. That the effect of 10$ spent in USA is equal to 10$ in china, so that they don't have to take money circulation in their model.
I find that assumption very false. 10$ that go to your local country is largely spent as tax to the country, rent, food, and healthcare services in the local economy. If you send 10$ to China, they won't be spending it on American restaurants, housing, or healthcare. Clearly USA should have some preference to say 10.1$ being spent in USA vs 10$ in china.
It is fairly easy to see what I mean, and why "free trade is always good" is only true for the whole system, not individuals or groups of individuals (countries).
To some extent the dollar is a reserve currency; people are willing to use it in place of their own currency so in some sense it will never come back to the U.S. To the degree this happens it's the equivalent of people taking paper from the U.S. and giving goods in return - which is to say, it's hard to understand how the U.S. loses on that deal.
If you still want to consider only usd and USA and China trade. Then usd is not just for purchasing us stuff. It is also a reserve currency that china holds.
It wasn't a question of currency at all, any wealth/money sent in a different economy is a loss to the original economy. A sizable gain in efficiency can offset that, but not every gain in free trade is of the required size, which is why tariffs should be desirable.
> I think protectionism and tariffs cannot be cast as inherently failures, and are necessary to build an economy[...]
I agree. There's a lot of cases to be made for bootstrapping an economy with central planning, protectionism and tariffs, e.g. post-WWII South Korea is a frequently cited example.
But this isn't the form tariffs tend to take in 1st world economies. They tend to take the form of concentrated benefits and higher diffuse costs to protect already failing industries that we mostly wouldn't have anymore if it wasn't for the tariffs propping them up.
Take two choices, you can either buy a 15$ tshirt in your local country, or a 10$ tshirt from another country. At an individual level it makes sense to buy the cheaper one.
But from a country's perspective, those 15$ spent in the economy are earned by someone in the economy to be spent again. On the other hand, the 10$ sent overseas have no effect of money circulation. Saving those 5$ over a t-shirt might cost 20$ in GDP. The calculations need to be done regarding velocity of money, taxation rate, and savings. But I do think that not all money is equal. 10$ spent overseas and 5$ spent domestically is less favorable than 15$ spent domestically.
I think protectionism and tariffs cannot be cast as inherently failures, and are necessary to build an economy.