In this note, we review a recent Federal Reserve note (“Detecting Tariff Effects on Consumer Prices in Real Time – Part II”) which has attracted considerable attention. As usual, we assess the authors’ methodology and provide our own critical evaluation. We also assess the implications for the Fed staff and the FOMC at the end. The key takeaway is as follows: the authors estimate that tariffs increased core PCE prices by up to 0.8% through February 2026. However, by construction, this figure should be interpreted as an upper bound. A more realistic estimate is likely lower, potentially closer to half of that magnitude.
What the paper does
The paper develops a framework to estimate the impact of tariffs on consumer prices, focusing on U.S. core PCE inflation during the 2025–2026 tariff episodes. The authors construct a measure of “theoretical tariff exposure” at the PCE category level, combining detailed information on tariff changes across products and countries with import shares and input-output linkages. They then estimate distributed-lag regressions of category-level inflation on tariff exposure, exploiting cross-sectional variation across goods categories. The empirical design includes time fixed effects, such that identification comes from differential inflation across categories with varying exposure to tariffs. The authors also provide an alternative specification without time fixed effects to capture aggregate effects. The analysis is conducted on a relatively short sample (January 2024 to February 2026), covering multiple overlapping tariff waves.
Main results
The paper finds that tariffs have a statistically and economically significant effect on consumer prices. More exposed categories experience systematically higher inflation, and the estimated pass-through is large and relatively fast, reaching close to 100 percent within several months. Aggregating across categories, the authors conclude that tariffs increased core PCE prices by up to 0.8 percent through February 2026 in their baseline specification. The authors argue that tariffs can account for most, if not all, of the excess core goods inflation relative to pre-pandemic benchmarks, and that pass-through is largely complete within a horizon of 5–9 months.
Figure 1. Tariff effects on core PCE prices
Our critical assessment
The use of cross-sectional variation in tariff exposure is a natural approach in a setting with multiple staggered tariff changes. The evidence that tariffs increased the prices of more exposed goods categories is convincing and aligns with prior intuition. However, the quantitative conclusions—most notably the 0.8 percent effect on core PCE—are not sufficiently robust to warrant the strong claims made in the paper. Indeed, the authors themselves caution against their results, a point that appears to have been overlooked by many.
First, the results are highly sensitive to the treatment of time effects. The baseline specification with time fixed effects identifies only relative price changes across categories, not aggregate inflation effects. Once time fixed effects are removed, the estimated impact on core PCE falls by roughly half, to about 0.4 percent. This is not a minor robustness issue but a central identification concern. It implies that the headline estimate is not stable across reasonable specifications and should not be interpreted as a precise measure of the aggregate inflation impact. (The intuition is the following: the time fixed effect absorbs all common shocks across categories, therefore mechanically magnifying the effects of the tariffs).
Second, the mapping from relative price effects to aggregate core PCE inflation is not fully justified. By construction, the baseline specification excludes common shocks affecting all categories simultaneously. Yet the paper interprets the results as capturing total inflation effects. This conflation between cross-sectional identification and aggregate inference weakens the credibility of the headline number.
Third, the claim that pass-through is “effectively complete” appears too strong given the short sample. Several tariff waves occur late in the sample, leaving insufficient time to observe full price adjustment. The conclusion relies in part on extrapolation based on assumed lag structures rather than directly observed outcomes.
Overall, while the qualitative conclusion that tariffs contributed to higher inflation is credible, the magnitude of the effect is highly uncertain. The evidence suggests a plausible range closer to 0.4–0.8 percent, rather than a precise estimate of 0.8 percent. In this sense, the headline figure is best interpreted as an upper-bound estimate and is likely somewhat exaggerated. Indeed, the authors write “if the smaller tariff effects obtained under the alternative specification (red line) are closer to the true effects, then other factors must have also contributed to keeping core goods inflation above its pre-pandemic levels in 2025. Determining the nature and persistence of these additional factors would then be relevant for the core goods inflation outlook, but such an analysis is beyond the scope of this note.”
The Fed staff and the FOMC
In our view, it is evident that this Fed note reflects the perspective of the Fed staff and, by extension, has likely influenced the FOMC as well. We believe this is one of the reasons—together with the assumptions regarding pi*—why both the staff and the FOMC continue to expect inflation to return to target relatively quickly. (Please, keep in mind that the Fed note says nothing about the persistency of core services) From our side, we can only reiterate that the last time core PCE inflation (year-over-year) was at target was 60 months ago, and that the current economic environment appears very different from the pre-Covid period. Unfortunately, over the past six years, the data have not supported the Fed’s view.