NEW YORK (TNS) — On hearing of the Continental Army’s pivotal victory at the Battle of Saratoga in 1777, John Sinclair told Adam Smith, “The British nation must be ruined.” As Sinclair recalled, the author of The Wealth of Nations (published the year before) urged him to calm down. “Be assured, my young friend, there is a great deal of ruin in a nation.”

Dedicated though he was to the benefits of free trade, Smith would doubtless say the same about today’s turn toward mercantilism. It’s a blow, but not the end of the world. That’s worth noting: Catastrophism, a popular mode of discourse these days, is usually unhelpful. But champions of President Donald Trump’s approach to trade are apt to make the opposite mistake — namely, thinking that if the roof hasn’t fallen in, the policy must be succeeding. If it results in slower growth and persistent under-performance, that might not be “ruin,” but it sure isn’t victory.

Once Trump’s new system of tariffs has settled down — if it ever does — what might it cost? What might “less than ruin” amount to?

According to most estimates, the direct economic losses are certainly tolerable, especially for a huge and relatively closed economy like the U.S. One recent study explores the upper limit on what’s at stake by calculating the benefits of liberal trade compared with no trade at all. For the U.S., the costs of closing the economy altogether would fall in the range of 2% to 8% of gross domestic product.

The costs of less trade, as opposed to no trade, would naturally be smaller still. Earlier this month the Federal Reserve published a research note on the effects of specific tariffs. Its economists modeled an increase of 60 percentage points in the U.S. tariff on imports from China, with and without a “baseline” tariff of 10% on other trading partners, assuming for one set of scenarios that the trade deficit is unchanged and for another that it shrinks.

According to their model, the 60% extra tariff on China, the 10% baseline tariff on everybody else plus a 25% reduction in the trade deficit would cut U.S. GDP by a little under 3%. (China’s losses would be about the same; thanks to shifts in the pattern of trade, the rest of the world would come out about even.)

These and other such studies reveal the complexity of the changes caused by trade barriers. For example, surely tariffs would reduce imports and hence shrink the trade deficit. Why assume, as some of the Fed’s scenarios do, that the deficit doesn’t change? Actually, it’s far from obvious that the trade deficit will narrow. You’d expect a smaller trade deficit to make the dollar appreciate — in due course increasing imports, cutting exports and undoing the initial effect. In any case, the overall external balance is determined by the gap between its saving and investment, which tariffs affect only indirectly.

Or consider the surprisingly small estimated cost of closing the economy completely. One of the assumptions behind the estimated losses of 2% to 8% of GDP is that the ease of replacing domestic goods with imports — the so-called elasticity of substitution — can be estimated from current trade data. But as the economy approaches autarky, this elasticity might fall abruptly as certain critical foreign products prove difficult or impossible to replace. The costs of abolishing imports might then be much bigger than projected. (Granted, a rational mercantilist would be careful not to press too far: An entirely closed economy isn’t the goal.)

The list of other complications is endless. What’s the effect of trade on competition and innovation? It depends. Up to a point, competition through trade is likely to spur innovation, but if foreign competition is severe enough to shut a domestic industry down, said industry won’t be more innovative. The dynamic effects of trade — that is, the effects of trade on growth — are even harder to estimate than the static effects captured in the studies mentioned above.

Amid all the uncertainty, two points seem worth emphasizing. First, despite the complexities, economists generally agree that trade does deliver net gains — that, on this, Adam Smith was right. If suppressing trade is costly, then exactly how costly is not the most important question. You don’t do it. To be sure, the U.S. has a huge domestic market and is richly endowed with natural resources. These advantages mean that trade is likely to deliver smaller gains than it does for other economies. But, to repeat, small gains are better than none.

Second, the costs of the new mercantilism aren’t confined to the implications for GDP of moving from a settled regime of liberal trade to a settled regime of managed trade. That shift involves massive economic and geopolitical dislocations, which are likely to be costly in themselves.

Economic restructuring expends resources; it creates jobs and destroys them. The “China Shock” was disruptive — but vainly trying to reverse it will be disruptive all over again. In the first case, there were aggregate benefits; in the second, there’ll be aggregate losses.

Geopolitical dislocation could involve the biggest costs of all. The new mercantilism puts U.S.-led alliances and multilateral institutions under enormous strain. The view that the U.S. has been exploited by these arrangements isn’t unwarranted — there’s been some free-riding, no doubt — but on balance U.S. global leadership has been an exercise in enlightened self-interest.

Dismantling the global trading order, and casting this as overdue retaliation against selfish so-called friends, is to cast away American power. It would be bad policy if undertaken in return for small economic gains. In return for substantial, even if less-than-ruinous, economic losses, it’s insane.

(Clive Crook is a Bloomberg Opinion columnist and member of the editorial board covering economics. He is a former deputy editor of the Economist and chief Washington commentator for the Financial Times.)