Tariffs Added 2.9 Percentage Points to Goods Inflation at Their Peak. The New York Fed Says That Effect Is About to Climb Again.

2.9 percentage points. That's how much of the rise in consumer goods inflation traces back to tariffs at its peak in February 2026, according to new research from the Federal Reserve Bank of New York. The analysis, published this past week by economists Mary Amiti, Sebastian Heise, and David Weinstein, is the most direct answer yet to a question that has dogged the Fed all year: how much of the inflation it is fighting was created by Washington's own trade policy.

The researchers tracked prices across dozens of consumer goods categories from 2025 into 2026 and compared what actually happened to a baseline of what prices would have done without tariffs. Their conclusion: goods prices would have drifted slightly lower over that stretch. Instead, tariffs pushed them up by 2.9 points at the February peak before the effect eased to roughly 2 points by August.

[[IMG1]]

What Changed

Two mechanics drive the gap between those numbers, and the researchers separated them cleanly. Import prices react almost immediately. About 90% of a given tariff increase shows up at the border within weeks. Retail prices move on a longer clock. Roughly half of the pass-through appears within three months, with the rest building over the rest of a year as importers work through existing inventory before repricing shelves and knock-on costs work their way through.

A third force adds to both: knock-on costs. Manufacturers that rely on tariffed parts and materials raise their own prices even when the finished product carries no tariff itself, and the researchers estimate roughly a third of the total effect comes from exactly that channel rather than the direct tariff line. On average, each one-point increase in the overall tariff rate lifts consumer goods prices by about a quarter of a percent after a full year.

The February-to-August easing has a specific cause, not a general cooling. The Supreme Court struck down a large share of the tariffs imposed under emergency economic powers that same month, and they were replaced with a flatter 10% import tax. That swap lowered the average rate on a wide swath of goods even as other, narrower tariffs on specific categories stayed in place or were added.

Why It Matters

The timing lands close to the next test of this exact question. September's Consumer Price Index report is due Wednesday, and Reuters' survey of economists points to a 3.7% headline reading and 2.5% at the core, which excludes food and energy. Treasury yields, already at levels last seen in 2000, and the Fed's own hike odds for December have been trading off exactly this kind of print all month, so a number that surprises in either direction carries outsized weight this week.

The New York Fed researchers do not think the tariff story is finished easing, either. Canadian levies are still working their way through supply chains, and a new round of auto parts tariffs is scheduled to take effect next year. Both, the authors write, should push the tariff contribution back up rather than let it keep fading toward zero.

That outlook complicates a Fed that just raised rates for the first time since 2023 and is weighing whether to go again in December. A separate New York Fed gauge, the Survey of Consumer Expectations, put one-year inflation expectations at 3.9% in September, the highest since May 2023, a different poll than the University of Michigan's but pointing the same direction. If households and researchers increasingly agree that tariffs are keeping a floor under goods prices, the case for treating this inflation as temporary gets harder for the Fed to make out loud, even if committee members believe it privately.

Bond markets have already been pricing something like this. The Treasury's October 7 sale of 10-year notes cleared at 5.300%, the highest since November 2000, even as foreign buyers absorbed 80% of the offering. That combination, a high yield with strong demand, suggests bond investors already believe current rates reflect real growth and inflation pressure, tariffs included, rather than a market mispricing risk.

What to Watch Next

Wednesday's CPI print is the immediate marker. A core reading above the 2.5% consensus would support the researchers' call for the tariff effect to climb again and likely firms up December hike odds further. A softer number would suggest the disinflationary forces elsewhere in the economy, like slowing hiring and a cooling housing market, are outweighing tariffs for now. The research team has said it plans to keep updating this estimate as new tariff rounds phase in, which gives this a concrete number to check back against rather than a one-time read.

The Pulse24 Take

Markets have spent over a year arguing about how much of this inflation cycle belongs to tariffs and how much belongs to everything else, a tight labor market, a housing shortage, a Fed that moved late. The New York Fed's answer is not a verdict on the whole debate, but it is a real number where there used to be a shrug. Roughly one of the few percentage points separating the Fed's 2% target from where goods inflation now sits can be traced to a policy choice rather than the business cycle, and Washington controls that lever directly. Wednesday's CPI report will not settle whether the Fed treats that distinction as decisive. It will show whether the tariff effect, fading since February, has actually reversed the way the researchers expect.