Who Actually Pays for Tariffs? What the Economics Research Says

A note on neutrality. Tariffs are politically charged; this piece is not. Gecko takes no position for or against any party, administration, or trade policy. Below is a fair summary of what the peer-reviewed economics literature actually finds — including the strongest arguments on more than one side — followed by what it means for a trader’s behavior. Nothing here is a policy endorsement.

Few economic questions are argued more loudly and understood less clearly than tariffs. The political debate runs on slogans — tariffs punish foreign countries, or tariffs are a tax on your own citizens — and both sides speak with total confidence. The useful thing about economics is that the 2018-19 US-China trade war was, in effect, a large natural experiment, and researchers have since measured what actually happened with unusual precision. The answers are clearer than the shouting suggests, and also more nuanced than either slogan allows. Here is what the literature really shows.

Key takeaways

• The best studies of the 2018-19 trade war found near-complete pass-through: US importers and consumers, not foreign exporters, paid essentially the full cost of the tariffs.
• Estimated losses ran to ~$1.4 billion a month of US real income (Amiti et al.) and ~$51 billion to US buyers, with a net welfare loss of ~$7.2 billion (Fajgelbaum et al.).
• The serious counterweight is the China Shock: Autor, Dorn & Hanson showed free trade inflicted deep, concentrated job losses — the diffuse-gains, concentrated-losses problem is real.
• For traders, the takeaway isn’t a policy view. It’s that tariff headlines are a volatility machine, and the research helps you understand the mechanism instead of trading the noise.

What the trade-war studies found: US buyers paid

Start with the most direct question: when the US put tariffs on imports, whose prices went up? In theory a tariff’s cost could be split — foreign exporters might cut their prices to keep customers, absorbing part of the tax. In practice, they didn’t. In “The Impact of the 2018 Tariffs on Prices and Welfare” (Journal of Economic Perspectives, 2019), Mary Amiti, Stephen Redding, and David Weinstein found complete pass-through: the duty-inclusive prices of imported goods rose by essentially the full amount of the tariff, meaning the entire burden fell on US importers and consumers. They estimated the tariffs were reducing US real income by about $1.4 billion per month by the end of 2018, before counting the damage from retaliation.

A second landmark study reached the same conclusion by a different route. In “The Return to Protectionism” (Quarterly Journal of Economics, 2020), Pablo Fajgelbaum, Pinelopi Goldberg, Patrick Kennedy, and Amit Khandelwal again found near-complete pass-through, and put the cost to US consumers and import-buying firms at about $51 billion (0.27% of GDP). After adding back the tariff revenue the government collected and the gains to the domestic producers who were now shielded from competition, the net aggregate loss was about $7.2 billion (0.04% of GDP). Smaller once you net it out — but a loss, not a gain, and concentrated in ways that matter: their analysis found workers in heavily agricultural and Republican-leaning counties were among the hardest hit, because those were the sectors foreign governments chose to retaliate against.

On the narrow question of who paid, the evidence is about as clear as economics gets: the tariffs were a tax, and Americans paid it.

The serious counterweight: the China Shock

If the essay stopped there it would be doing exactly what the sloganeers do — presenting one true thing as the whole truth. So here is the other side, and it is not a talking point; it is some of the most influential economics of the last two decades. In a series of papers beginning with “The China Syndrome” (2013), David Autor, David Dorn, and Gordon Hanson documented that the flood of Chinese imports after 2000 caused large and persistent job losses in the specific US local labor markets most exposed to that competition — on the order of one to two million jobs across the economy — and, crucially, that the affected workers and towns did not smoothly retrain and relocate the way textbook trade theory assumed. The pain was deep, local, and long-lasting.

This is the intellectual foundation beneath the modern turn toward protectionism, and honest economics has to sit with it. The gains from free trade are real, but they are diffuse — slightly cheaper goods for everyone — while the losses are concentrated on particular workers and communities who rarely receive the compensation that theory promises. A trade policy can be net-positive for the country and still devastate a region, and telling that region the aggregate math works out is neither persuasive nor, arguably, just. The China Shock authors documented this cost; notably, they did not conclude that broad tariffs are the efficient remedy. But they gave the protectionist impulse an evidence base it had previously lacked.

The arguments, laid side by side

Beyond the two headline findings, economics recognizes a set of genuine arguments on each side. Presenting them fairly is the point.

Where does the profession net out? On the broad, across-the-board use of tariffs, the mainstream is decidedly skeptical: surveys like the University of Chicago’s IGM Forum, and older polls of PhD economists, show overwhelming agreement that tariffs and quotas usually reduce overall economic welfare. But “usually reduce welfare” is a statement about the average case, not a denial that targeted, strategic, or security-motivated exceptions can be defensible. The honest summary is that economists broadly oppose broad tariffs on efficiency grounds, while acknowledging real distributional and strategic considerations that pure efficiency arguments miss.

Why this belongs on a trading blog

Because you don’t trade the policy — you trade the market’s reaction to it, and those are very different things. The 2025 tariff episode we covered in the headline-volatility essay is the perfect illustration: the “Liberation Day” announcement triggered the largest two-day loss in market history, then a near-complete recovery within weeks once a pause was announced. The economics above unfolds over quarters and years — prices, supply chains, jobs, retaliation. The market repriced most of it in days, and then repriced the reversal in days more. A trader who understood the research perfectly and traded the headline emotionally still got whipsawed.

That’s the behavioral lesson. The research is genuinely useful for understanding the mechanism — who bears the cost, which sectors win or lose, where retaliation lands — and that understanding can inform a patient, well-sized view. What it cannot do is give you an edge in the frantic hours after an announcement, when the move has usually already happened and the only thing being tested is your discipline. As with the Fed and geopolitical shocks, the loud, scheduled-feeling macro event is where behavioral errors are punished hardest, not where retail edge is found.

Knowing what tariffs really do is worth a great deal. Trading the tariff headline as if that knowledge is an edge is worth rather less — and usually costs more.

From belief to behavior: don’t trade the macro narrative

Resources and further reading

• Pass-through & welfare: Amiti, M., Redding, S. & Weinstein, D. (2019), “The Impact of the 2018 Tariffs on Prices and Welfare,” Journal of Economic Perspectives 33(4): 187-210.
• The trade-war cost: Fajgelbaum, P., Goldberg, P., Kennedy, P. & Khandelwal, A. (2020), “The Return to Protectionism,” Quarterly Journal of Economics 135(1): 1-55.
• The counterweight: Autor, D., Dorn, D. & Hanson, G. (2013), “The China Syndrome,” American Economic Review; and Autor, Dorn & Hanson (2016), “The China Shock,” Annual Review of Economics.
• The consensus: University of Chicago Booth IGM Forum surveys of economists on tariffs and trade.
• The history: Douglas Irwin’s work on Smoot-Hawley (1930) and the economic history of US trade policy (Peddling Protectionism; Clashing over Commerce).

Frequently asked questions

Who actually pays for tariffs, according to economists?

The best-identified studies of the 2018-19 tariffs found near-complete pass-through: import prices rose by roughly the full tariff, so US importers and consumers bore essentially all the cost. Amiti et al. estimated ~$1.4B/month of lost US real income by late 2018; Fajgelbaum et al. estimated ~$51B in losses to US buyers and a ~$7.2B net welfare loss after tariff revenue and producer gains.

Do economists support tariffs?

On standard welfare grounds, a large majority are skeptical of broad tariffs — IGM and older surveys show overwhelming agreement they usually reduce overall welfare. But economics also recognizes exceptions: national security, strategic/infant industries, terms-of-trade effects for large countries, and the concentrated harms free trade can cause. “Economists oppose broad tariffs” is the mainstream view, not an unconditional one.

What is the China Shock and why does it matter?

Autor, Dorn & Hanson found the post-2000 surge in Chinese imports caused deep, persistent job losses (roughly 1-2 million) in the US local labor markets most exposed, which adjusted far more slowly than theory assumed. It’s the evidence-based counterweight to “free trade is costless”: trade’s gains are diffuse, its losses concentrated. The authors documented that cost — which is distinct from endorsing tariffs as the fix.

What do tariffs mean for traders?

Mainly volatility. The research clarifies the real mechanism (higher prices, sector winners/losers, retaliation), but the dominant near-term effect is headline-driven whipsaw, as 2025 showed. The disciplined response is regime awareness — smaller size around announcements, no reacting in the first minutes — not a bet on trade politics.

Essay in Gecko’s trading psychology series, and explicitly non-partisan: it summarizes peer-reviewed findings and their strongest counterarguments, not the merits of any party or policy. Figures are drawn from the cited studies (Amiti, Redding & Weinstein 2019; Fajgelbaum, Goldberg, Kennedy & Khandelwal 2020; Autor, Dorn & Hanson) and are specific to the periods and methods studied; other studies reach somewhat different estimates. Gecko is an educational and informational tool. Nothing here is financial, investment, or trading advice, a policy endorsement, or a political statement. Trading carries substantial risk of loss.