In this note, we explain what happened to the supply side of FRB-US. Part I of the note is here. The bottom line is that according to the model, the US economy is stronger than previously estimated, as FRB-US revised up both the level and the slope of potential output (Y*). As such, the model projects the US economy to remain strong and the unemployment rate to fall in the second half of next year.
The updated supply side
Higher level, higher slope of Y*. Figure 1 shows the level (left panel) and the slope (right panel) of estimated potential output in FRB-US (SEP-consistent database). The blue line is the latest dataset, while the gray line is the dataset at the time of the July FOMC. The bottom line is that both the level and the slope of Y* are higher, partially because of the BEA annual revision, but partially because the model estimates a stronger, more productive economy. (For the record, trend participation and civilian employment in the forecast are both little changed in the September database, therefore the upward revision in the left panel of Figure 1 comes mainly from the BEA revision) The big news is that potential growth is estimated close to 2½ percent in 2022-2024, although it gradually converges back to 2% by the end of the medium-term.
Figure 1. Level (left panel) and slope (right panel) of potential output (Y*)
Output gap: positive or negative?
Because the model revised up the level of Y* (and more than the upward revision of Y from the data), the output gap is revised down. Figure 2 shows the level of real GDP and potential output (left panel – solid blue line and dashed green line, respectively) in the latest dataset, and a comparison of the output gap July-September (right panel). The model was estimating a small negative output gap in July; the model continues to estimate a small negative gap in 2024 now but projects it forward until 2027. By looking at Figure 2 and thinking about the Okun’s law (as well as about other “gaps”), our impression is that the model is overstating, at least a bit, the level of Y* in 2024. Therefore, the impression is that the output gap is possibly smaller than in Figure 2. In any case, to us, the message comes from the slope of Y*: the model is sending the message that the economy remains quite strong and can keep growing at 2%+ going forward.
Figure 2. Real GDP (solid blue), potential output (dashed green), and the output gap
Model uncertainty
Model uncertainty is high, output gap is likely small and positive. While we use FRB-US extensively, including for simulations, we take any single model’s estimates with a grain of salt. Figure 3 shows a comparison of estimated output gaps. As visible, the dispersion is very high, with some models suggesting a positive but decreasing gap, and FRB-US suggesting a small negative gap. The average across models is around 1%, and it is projected to remain around the current level going forward. Does this imply that the FRB-US estimates are “wrong”? No, as we explain in the conclusion.
Figure 3. Estimates of the output gap – 6 models and average across models
Conclusion
We are dealing with a strong economy, hard to see a recession without a shock. As we will show in Part III of this note, it remains hard to see a recession without a shock (please, note that this is the same result we got in the last 4 quarters or more…). According to FRB-US, potential is higher, and the economy is more productive. But even if the reader considers the estimated level (by FRB-US) of Y* too high -which, again, is possible by looking at Figure 2 or thinking about the Okun’s law- other models continue to estimate a positive output gap in 2024. Bottom line is that the economy seems fine (in fact, even stronger). A recession is always possible, but our guess is that we need another shock before it will happen.