The Anti-Tariff Priesthood Strikes Again

NBER Study an Indictment of Academic Class Captive to Free-Trade Ideology

Team,

This is my new piece dismantling a recent paper published by the National Bureau of Economic Research (NBER). The study purports to expose the folly of Trump’s trade strategy, but its own findings undermine its anti-tariff narrative.

You can also read my piece HERE in American Greatness.

Far from being an indictment of Trump’s tariffs, the NBER study is an indictment of an academic class so captive to free-trade ideology and so boxed in by its own tools that it cannot model tariffs as they actually work in the real world. Its own evidence shows that tariffs raised revenue, hit China hard, and reshaped sourcing patterns in ways the authors’ static welfare accounting is poorly equipped to value. In the end, it is just the latest entry in a long tradition of flawed tariff analysis.

I always appreciate your comments. Please be sure to share this.

Peter

The Anti-Tariff Priesthood Strikes Again

When a model swings from loss to gain on its own assumptions, it isn’t proving tariffs failed—it’s exposing the limits of the model.

By Peter Navarro

April 16, 2026

The anti-tariff priesthood has delivered yet another paper purporting to expose the folly of Trump’s trade strategy. Exhibit A is the latest National Bureau of Economic Research (NBER) study from Pablo Fajgelbaum and Amit Khandelwal, which tells us that Trump’s 2025 tariffs pushed U.S. duties from 2.4 percent to 9.6 percent, that about 90 percent of those tariffs were passed through into prices paid by U.S. importers, and that the net welfare effect ranged from a 0.13 percent loss of GDP to a 0.10 percent gain.

That is the tell. When your headline conclusion runs from minus to plus depending on assumptions the authors themselves admit are unresolved, you do not have a verdict. You have a shrug dressed up as science.

The paper offers a few careful concessions, just enough to establish an air of fairness. Yes, the tariff picture was more complicated than the daily media hysteria suggested. Applied tariffs were often lower than statutory headline rates, a majority of imports still entered duty-free, and retaliation outside China was limited.

Those are important facts, and they cut against much of the panic-porn that dominated coverage in 2025. But once those concessions are out of the way, the authors do what tariff critics always do: they force a strategic policy through a static model built to find short-run distortion and miss long-run gain.

The paper expressly says its framework is static, has exogenous trade deficits, and lacks monetary frictions and uncertainty. In other words, it is built to capture short-run price distortions, not strategic industrial realignment.

That matters because the Trump tariff program was never mainly about squeezing one more decimal point of static consumption welfare out of a one-period trade model. It was, and is, about revenue, bargaining leverage, supply-chain resilience, reducing strategic dependence on China, and stopping not just China but much of the rest of the world from cheating America.

And here is where the paper’s own findings undermine its anti-tariff narrative. The authors concede that the tariffs succeeded in generating substantial federal revenue and accelerating decoupling from China. Tariff revenue rose to 4.9 percent of federal receipts, up from 1.6 percent in the prior decade. China’s share of U.S. imports fell to 7 percent in December 2025, down from 23 percent in December 2017. Those are not rounding errors. Those are major strategic shifts.

Now consider pass-through, the favorite weapon in every tariff-bashing paper. In this study, the headline number is 90 percent: in the authors’ baseline quarterly specification, about 90 percent of the tariff shows up in prices paid by U.S. importers—an estimate that leaves little room for foreign exporters to bear much of the burden. That sits uneasily with evidence from the first Trump tariff episode, when foreign producers absorbed a meaningful share of the burden through lower prices, weaker margins, and currency adjustment.

Here, as I explained in The Wall Street Journal, the company that writes the check to Customs is not necessarily the party that ultimately bears the economic burden of the tariff. That burden gets worked out through a much broader adjustment process—exporter price concessions, margin compression, quantity reductions, supplier switching, and currency moves. So even on its own terms, what this paper is measuring is not the final burden of the tariff economy-wide. It is short-run pass-through at the border.

The paper itself spends several pages explaining how messy tariff measurement really is. Statutory and applied rates diverge because of mid-month changes, shipping lags, bonded warehouses, Section 232 input-origin complications, USMCA-compliant entries, and tariff exclusions.

Yet when the authors turn to pass-through, they use statutory tariffs as a proxy for applied tariffs even while conceding that the relationship between the two weakened as those measures diverged in 2025. They also use product-time and origin fixed effects, even while acknowledging that those controls can absorb front-running and broader general-equilibrium adjustment. That is not econometric precision. It is econometric murk.

The paper itself notes that pass-through at the border is not the same as the burden on final consumers, because retail incidence depends on distribution margins and on how much of the final price is actually imported content.

So the paper’s alleged smoking gun—the 90 percent pass-through number—is not a hard fact about who paid for the Trump tariffs. It is a model-dependent snapshot built on a quarterly window, a particular choice of fixed effects, and a statutory-to-applied mapping that the authors themselves say is imperfect. That is the same mistake virtually every tariff-bashing analyst makes: confusing a narrow border-price regression with the full economic incidence of tariffs.

The same overreach infects the welfare discussion. The NBER paper says the welfare effect is either slightly negative or slightly positive, depending on whether terms-of-trade effects show up. But the authors also say the data are inconclusive on precisely that question and that more work is needed to identify those effects.

So again, the central policy conclusion depends on the very mechanism that the paper cannot pin down. Worse still, once labor mobility is allowed, the range widens further. That is not a robust indictment of Trump’s tariffs. It is evidence that the model is too narrow and fragile to carry the rhetorical weight the anti-tariff crowd wants to place on it.

Then comes the biggest conceptual mistake: confusing short-run trade arithmetic with long-run strategic success. The authors themselves acknowledge that their framework omits capital accumulation, international capital mobility, intertemporal substitution, imperfect competition, and increasing returns. As John McEnroe might say, “You cannot be serious.”

They also admit that market-access gains and foreign-investment commitments generated by bilateral bargaining are beyond the scope of the paper. That is a striking omission, because one of the greatest contributions of the Trump tariffs has been precisely the leverage they created—and the deals with the U.K., Japan, the E.U., Vietnam, South Korea, and others that leverage made possible.

They also point to weak short-run manufacturing job evidence. But nobody who understands industrial reorganization believes a tariff wall goes up in April and a new factory magically opens in May. Reshoring is an investment process, not an event-study chart. The paper itself notes that more time is needed to judge those objectives. No kidding.

So what is the real takeaway? Far from being an indictment of Trump’s tariffs, this NBER study is an indictment of an academic class so captive to free-trade ideology and so boxed in by its own tools that it cannot model tariffs as they actually work in the real world. Its own evidence shows that tariffs raised revenue, hit China hard, and reshaped sourcing patterns in ways the authors’ static welfare accounting is poorly equipped to value. In the end, it is just the latest entry in a long tradition of flawed tariff analysis, joining the earlier work of Fajgelbaum, Goldberg, Kennedy, and Khandelwal; Amiti, Redding, and Weinstein; and Cavallo, Gopinath, Neiman, and Tang.

* * *

Peter Navarro is the White House Senior Counselor for Trade and Manufacturing. His website is www.peternavarro.com.