US: A Structurally More Dovish Fed

The Staff Forecast

The minutes include two key passages:

  • “Tariff increases were still expected to raise inflation this year and to provide some further upward pressure on inflation in 2026. Inflation was projected to decline in 2026, to reach 2 percent in 2027, and to remain there in 2028.”
  • “The staff continued to expect that the labor market would soften further this year, though the projected path for the unemployment rate in following years was a little lower than in the previous staff forecast. The unemployment rate was projected to move slightly above the staff’s estimate of its natural rate through the remainder of this year and then to decline later in the projection as GDP growth picked up.”

Translation. According to the staff: (1) There is little or no slack in the labor market this year and going forward, (2) Trend inflation (Pi*) is assumed to be at 2%, and (3) The tariff shock is treated as temporary and non-persistent.

The staff (core) inflation forecast

For readers unfamiliar with the mechanics of the Fed staff forecast, we refer to our 2022 note (here) and our 2025 video (here). Figure 1 illustrates our decomposition of the staff forecast, while Table 1 reports the corresponding values. For simplicity, one can interpret the decomposition as the contribution of various factors estimated through an augmented Phillips curve.

Figure 1. The Fed staff forecast according to our reverse engineering.

Note: The above shows our reverse engineering of the Fed staff forecast. The above shows the judgmental contributions informed by an estimated augmented Phillips curve model. The assumption about pi* flat at 2% in the forecast is inferred from Peneva et al. (2025).

Table 1. The Fed staff forecast according to our reverse engineering.

Note: The above shows our reverse engineering of the Fed staff forecast. The above shows the judgmental contributions informed by an estimated augmented Phillips curve model. The assumption about pi* flat at 2% in the forecast is inferred from Peneva et al. (2025).

Critical Points

As we have discussed in previous notes, the staff forecast is currently dovish, largely because of two key assumptions embedded in the decomposition.
The first is that pi*—or underlying inflation—is flat at 2 percent. This assumption mechanically drives the forecast to converge to target rather than remain above it.
The second is that the tariff shock is assumed to have no persistence—treated as a one-time, level effect on marginal costs rather than a slope effect on the Phillips curve.
Added to these is the assumption that the labor market exhibits no slack, meaning that the contribution from labor-market channels is effectively zero this year and beyond.

Our Assessment

In our view, both of these assumptions—on pi* and on tariffs—are quite generous. Our models indicate a high probability that underlying inflation is currently well above 2 percent and is likely to remain above target in the medium term. Moreover, the assumption that the tariff shock is purely transitory, rather than affecting the slope of the Phillips curve, is strong and may prove inconsistent with incoming data (the reader wants to take a moment to think about that the yellow bars in Figure 1 are non-zero mean and they are serially correlated..).

That said, the staff has made a clear strategic choice, and the FOMC’s core PCE forecast effectively mirrors the staff’s baseline, as is typically the case. By construction, these assumptions yield a distinctly dovish outcome. Based on our experience, it is highly unlikely that the staff will revise its pi* assumption or the broader forecast framework in the near term. Consequently, we conclude that the Fed has now moved toward a structurally more dovish stance—one that is likely to persist for months (or years).

Contact

Want something more tailored?

We provide tailored consulting on ad-hoc projects.