It’s no secret that one of Donald Trump’s favorite U.S. presidents is William McKinley, who led the country from 1897 until his assassination in 1901. Indeed, Trump recently changed the official name of Denali back to Mount McKinley in honor of the late president.

In recent months, Trump has particularly emphasized McKinley’s protectionist policies, which Trump believes were a key part of America’s economic progress in the McKinley era. “President McKinley made our country very rich through tariffs and through talent,” Trump said in his second inaugural address. “[H]e was a natural businessman—and gave Teddy Roosevelt the money for many of the great things he did.”

The idea that tariffs historically helped spur American prosperity has become a common refrain for many in the Republican Party. And at first blush, it certainly seems plausible—the McKinley era witnessed a fair amount of protectionism alongside considerable economic growth.

But there is a fatal flaw with this view, which has been pointed out most explicitly by Ludwig von Mises and his fellow Austrian economists. Understanding their perspective is critical if we want to craft good trade policy.

One of the important features of Austrian economics in the Misesian tradition is its approach to interpreting historical events. According to Mises, we can’t simply infer causation from the course of history, because too many factors are at play and it’s impossible to control for them all.

Consider the McKinley era. During this time technology was advancing rapidly, demographics were changing, transportation infrastructure was being built—lots and lots was changing. So simply noting that protectionism and economic progress happened side by side doesn’t really demonstrate that the former was any help to the latter. For all we know, the economy might have been progressing due to other factors in spite of the tariffs, and would have progressed even more if they had been lifted.

To understand which interpretation is correct—whether tariffs were propelling prosperity or putting a check on it—we need an a priori theory. It is only by developing sound economic laws independent of history that we can actually understand what is going on in any historical event. As Mises writes in his 1957 book, Theory and History:

“…historical experience is always the experience of complex phenomena, of the joint effects brought about by the operation of a multiplicity of elements. Such historical experience does not give the observer facts in the sense in which the natural sciences apply this term to the results obtained in laboratory experiments…Historical facts need to be interpreted on the ground of previously available theorems. They do not comment upon themselves.”

To infer causation simply by looking at history with no theorem is really to commit the post-hoc fallacy: A happened, then B happened, therefore A caused B. In our case: tariffs were implemented, then the economy prospered, therefore tariffs caused prosperity.

But this is a non-sequitur! One can hypothesize based on the facts that the tariffs were an aid to prosperity, but it’s also possible that other factors were behind the economic progress, and that the tariffs held it back, or even that they didn’t make any difference. We can’t know just by noting that these two events happened sequentially in history. And yet, that seems to be the essence of Trump’s argument—that this sequence of events somehow proves that tariffs were good for America.

Elaborating on the same point in his 1949 treatise Human Action, Mises cautions us against those who weaponize history to push a political agenda:

“Those writers who consider historical events as an arsenal of weapons for the conduct of their party feuds are not historians but propagandists and apologists. They are not eager to acquire knowledge but to justify the program of their parties…They usurp the name of history for their writings as a blind in order to deceive the credulous.”

If we can’t look to history to tell us whether tariffs are helpful or harmful, then how can we know whether they are a good idea? As Mises indicates, the way we can know is by consulting sound economic theory.

Mises summarizes what economic theory says on the matter of international trade later in Human Action:

“In this field the teachings of the classical economists, especially those of Ricardo, are final and settle the issue forever. All that a tariff can achieve is to divert production from those locations in which the output per unit of input is higher to locations in which it is lower. It does not increase production; it curtails it.”

Based on this economic understanding, we can confidently say that the tariffs in the McKinley era were—generally speaking—working against America’s prosperity. America would have progressed even more in their absence.

Now, some may be tempted to dispute this conclusion, and that’s fine. But there is a valid and an invalid way to dispute it. The valid way would be to question the reasoning of economist David Ricardo, to demonstrate some flaw in the logical system of economic theory. The invalid way would be simply pointing to some historical period where tariffs and economic growth happened to coincide and saying “See, that settles it, tariffs are good for the economy.”

It feels silly that this even needs to be said, and that Ludwig von Mises had to dedicate much of his career to explaining this simple point. But if powerful figures insist on justifying bad policies with bad historical arguments, someone needs to call them out on it.