President Trump thinks trade wars are “good and easy to win,” mostly, it seems, because he doesn’t understand anything about them.

Which, to be honest, isn’t really news. Trump, after all, has just been saying the same false things about trade the last few days that he has been saying the last few years. The only difference is that he has put so many tariffs in place in the meantime that, according to Deutsche Bank Chief Economist Torsten Slok, the United States will be taxing imports at a higher rate than do a lot of emerging market countries if Trump follows through on his threat to increase them on Friday.

It might have taken a few tweets, then, but the trade war is starting to get serious.

What exactly is it, though, that Trump gets wrong? The easiest way to think about it is that he almost comically overestimates how much trade hurts us, and just as absurdly underestimates how much tariffs do. The bedrock of these bad ideas is his belief that we “lose” money whenever we buy more goods and services from another country than it does from us. So, to take one particularly pertinent example, he thinks that our nearly $400 billion trade deficit with China means that, yes, we’re losing $400 billion to them (although, in true Trumpian fashion, he exaggerates this to be $500 billion instead). This is a strange sort of cost-benefit analysis that ignores the benefits of imports (what we get), and looks only at the costs of them (what we pay). It’s like saying that we lose money anytime we go shopping anywhere, because stores aren’t buying anything from us in return. Which, when you think about it, is almost a pre-money view of the world where being able to barter goods for each other is the most important measure of economic strength.

The result is that Trump doesn’t think we have much to lose from blowing up the global trading system, since, according to his extremely incorrect theory, that system is already making us lose hundreds of billions of dollars a year. It’s an optimism, in other words, born of ignorance. Which is why it almost goes without saying that the reality of a trade war would be much more pessimistic. There are two things to keep in mind here. The first is that there really are costs to trade, especially when other countries aren’t playing by the rules. In particular, it can make us lose good-paying manufacturing jobs when the economy is doing well and overall jobs when it’s not. Altogether, though, these costs are nowhere near the $500 billion annual ones that Trump likes to bandy about.

The second point is that the benefits of trade tend to be a lot larger than whatever it does cost us. That’s because, by allowing us to specialize in things we’re better at, trade increases our productivity in the same way that new technologies do. Throwing a big enough wrench — or tariff, as the case may be — in that system would slowly make us a little poorer each year. Not in a way that would be immediately obvious but, rather, in the sense that we’d all just be a little less well-off than we should be.

Trump isn’t worried about any of this, though, because he doesn’t understand how tariffs work, either. In his mind, they’re almost a form of tribute that other countries are forced to pay us. This, unlike his other ideas, at least has the virtue of being possibly true. But it just so happens that we know it’s not, this time around. How is that? Well, it all comes down to whether China, in this case, would rather try to maintain its market share or its profit margins. If it chooses the first, it will cut its prices to fully offset the impact of our tariffs — keeping the prices that U.S. consumers pay the same — which is basically the same thing as paying those tariffs itself. If it chooses the second, though, it won’t cut its prices at all, and will pass on the cost of these tariffs to its American customers instead ...

. . . Which is precisely what a team of researchers from Princeton, Columbia and the New York Federal Reserve found happened in the first year of Trump’s trade war. The simple story is that, because Chinese companies didn’t charge less, U.S. households ended up paying more. U.S. businesses did, too, but the U.S. government still didn’t get enough money from its tariffs to make up for these other losses. As a result, the country as a whole was slightly poorer.

The real problem, though, isn’t just that Trump is in thrall to a bad idea that works a lot worse than he thinks. No, it’s that Trump’s advisers might also be in thrall to it, because they think it’ll work a lot better for them, politically speaking, than anything else. That being seen to be fighting a trade war with China is preferable to not doing so, even if it’s economically self-defeating (which it, of course, is).

The irony, then, is that trade wars really might be good, as far as Trump is concerned, because of the fact that they aren’t easy to win.