New York Fed economists put the tariff lift to US consumer goods prices at 2.9 points

On the rate of inflation, the tariff effect has already turned. The researchers said annual consumer goods price inflation peaked early this year ^[9]. The tariff effect on the price level slid from near 3% in February to around 2% by August, Quartz reported ^[11]. That took roughly a third off the peak in six months ^[18]. The drop followed a Supreme Court ruling that invalidated the tariffs levied under emergency powers, and a lower 10% import surcharge replaced them ^[11].

On the price level, the effect stays. Goods prices remain above where they would have been without tariffs even as the inflation contribution fades ^[12], and the researchers estimate that without tariffs those prices would have dipped modestly ^[13]. I think the one-time-bump view is right about this year's goods inflation and wrong about the prices households pay. Those prices still carry about 2 points of tariff ^[11].

The 2027 part of the forecast rests partly on a tariff that is not yet in force. The researchers expect the tariff contribution to twelve-month goods inflation to turn slightly positive again by mid-2027, driven by tariffs on Canadian goods already in place and a planned increase on Canadian automobiles in January 2027 ^[8]. Tariffs reach import prices almost immediately. Effects on US-made goods take nine to twelve months to move through supply chains ^[7]. On that lag, the domestic part of the January auto increase would reach consumers between October 2027 and January 2028 ^[17].

The pass-through numbers measure different things. Nearly 90% of a tariff shows up in import prices, so foreign exporters absorb very little of it ^[5]. Across all three channels, about 26% of a tariff increase reaches the consumer goods price level ^[4]. Put another way, each point on the average tariff rate adds about 0.25% to goods prices a year later ^[3]. Two-thirds of the effect came directly through imported goods ^[6], about 1.9 of the 2.9 points ^[14]. The remaining point or so came from US producers paying more for imported inputs and from domestic firms raising prices against weaker foreign competition ^[6].

Divide 2.9 by the 26% pass-through and the figures imply an average tariff-rate rise of about 11 points ^[15]. Or rather, at least 11, because part of February's effect on US-made goods may still have been working through the nine-to-twelve-month lag ^[7]. The published summaries do not report the tariff rate itself.

The January auto increase is the test. If it lands and the tariff contribution to goods inflation turns positive by mid-2027, the forecast holds ^[8]. If it lands and the contribution is still at or below zero a year later, the pass-through estimates are overstating what reaches consumers. Should the increase be dropped, the forecast is left with only the Canadian tariffs already in place as a driver ^[8]. The paper, first published in August and revised in September ^[10], is built on 67 non-oil goods categories and excludes services ^[2].