Is Figura and Waller (2022) Optimistic? Yes, is the Answer

Keep in mind

Andrew Figura (Fed Board officer) and Chris Waller (Fed Board Governor) have recently published a FEDS note claiming that a large drop in the vacancy rate can occur without a sizable increase in the unemployment rate.

Taking Figura and Waller’s scenario as given, our research shows that: (i) their scenario is barely enough to disinflate the US economy without help from the supply side, and (ii) other measures of jobs opening and labor turnover produce a forecast of core PCE price inflation well above the Fed target in 2024.

For this reason, in our view, Figura and Waller (2022) should be taken as a best-case scenario, not as a baseline.

Figura and Waller (2022)

Andrew Figura and Chris Waller have recently published a FEDS note titled “What does the Beveridge curve tell us about the likelihood of a soft landing?”. The note is particularly relevant because of the authors: one (Waller) is a Governor and the other (Figura) is a Fed Board officer supervising the labor market and the inflation desks. For this reason, Figura and Waller (2022) can be taken as proxy of the Fed staff view.

Figura and Waller (2022) is a theoretical analysis of the Beveridge curve investigating what conditions are necessary for a “soft landing” in the labor market. Without discussing the details, Figura and Waller (2022) use a Beveridge curve framework to estimate the standard parameters of a matching function. Then, they derive the implied change in the unemployment rate resulting from a decline in the vacancy rate. Figura and Waller (2022) claim that the change in unemployment rate resulting from a decline in the vacancy rate to 4.6 percent is small: about 1 percent. Therefore, Figura and Waller (2022) argue that a “soft landing” is, indeed, possible. The intuition of Figura and Waller (2022) is quite simple: the authors argue that the US economy is currently on the (very) steep part of the Beveridge curve. Consequently, vacancies can fall without triggering a large increase of the unemployment rate.

(Waller and Figura (2022) is a reply to Blanchard, Domash, and Summers (“Bad news for the Fed from the Beveridge space“) who argue that fighting inflation will require a reduction in jobs vacancies and also an increase in unemployment. Blanchard, Domash, and Summers have published a counter-reply to Figura and Waller (2022) titled “The Fed is wrong: Lower inflation is unlikely without raising unemployment”. In this debate we are agnostic. The reason is that the results are quite sensitive to the estimated parameters. Not only, but from our perspective there are other more important questions – see below)

The research questions

In their analysis, Figura and Waller (2022) define “soft landing” as a decline in the vacancy rate from 7 percent to 4.6 percent. However, Figura and Waller (2022) assume (and never discuss) that a decline in the vacancy rate to 4.6 percent is enough to disinflate the US economy. Is their assumption correct? Is it robust? What happens if rather than looking at the vacancy rate we use other labor market measures of slack, such as the ratio of vacancies to unemployed or the quits rate? These are the research questions we investigate in this note.

Assumptions and methods

Figure 1 shows the path of three measures of jobs opening and labor turnover: the jobs opening rate (the blue line), the ratio between openings and unemployed (the red line), and the quits rate (the yellow line). The solid lines show published data, while dotted lines show our assumptions going forward. Consistent with Figura and Waller (2022), we assume the jobs opening rate to return to 4.6 percent by the end of 2024, a level roughly in line with the pre-Covid period. As for the openings per unemployed and the quits rate, we also assume a “soft landing” (to 1.2 percent and 2.1 percent, respectively).

Given these assumptions, we use the three measures of jobs opening and labor turnover as alternative variables of slack in our “main” Phillips curve model to forecast core PCE price inflation (needless to say, each variable is a substitue of the others in the regressions). All other variables and parameters in the Phillips curve are unchanged. The three measures can either enter in level or in gaps (computed either using a statistical trend or the Z-score). Our results are not sensitive to this choice.

Figure 1. Measures of jobs opening and labor turnover

Note: the figure shows three measures of jobs opening and labor turnover. Published data are taken from the BLS and indicated by a solid line. Forecasts are indicated by dotted line.

Results

The upper part of table 1 shows our assumptions for jobs opening rate (line A.), openings per unemployed (line B.), and quits rate (line C.). The forecast horizon is 2022-2024. The bottom part of table 1 shows our model-based forecasts for core PCE price inflation using the three measures of slack. Figure 2 shows the forecast plots.

Three results emerge from table 1. First, under the assumptions of Figura and Waller (2022) core PCE price inflation barely converges back to 2 percent by 2024 (line 1). This forecast is remarkably close to the June SEP (and June staff forecast, as inferred from the FOMC minutes). Second, the forecasts using openings per unemployed (line 2) and quits rates (line 3) remain (well) above target at the end of the medium-term. Finally, the forecasts using openings per unemployed (line 2) and quits rates (line 3) are both above 3 percent in 2023.

What explains the differences in the forecasts?

The three measures of slack we consider are highly correlated (figure 1) but the trajectory of price inflation in the forecast is a bit different, especially for openings per unemployed. Our research suggests that jobs opening rate and quits rate are more correlated to core PCE price inflation than openings per unemployed. In other words, the Phillips curve is flatter in the openings per unemployed regressions. As such, even if all measures go back to their 2019 levels by the end of 2024, core PCE price inflation remains higher when using openings per unemployed because the flatness of the Phillips curve would require a lower level of slack (a more negative gap).

Which forecast is more reliable?

Our exercise is not meant to provide a best forecast. Instead, our analysis has two goals: (i) verify whether Figura and Waller (2022) “soft landing” claim is correct, and (ii) highlight the risks around their claims. As for (i), indeed our results show that lowering the jobs opening rate to 4.6 percent is barely enough to bring inflation close to target. Nevertheless, because JOLTS data start only in the 2000s, the size of the sample is limited. Therefore, all results should be interpreted carefully. Regarding (ii), our results show that the risks around Figura and Waller (2022) claim are definitely skewed to the upside. Indeed, the authors themselves recognize that while a drop of the jobs opening rate to 4.6 is large given the initial level, it would imply an “historically still quite high” end point (Figura and Waller (2022). Put it differently, in our view and estimates, the authors are considering a (quasi) “best case” scenario as a baseline. Consequently, it should not be surprising that the risks around their implied-forecast are skewed to the upside.

Table 1. Assumptions and core PCE price inflation forecasts

Conclusion

Figura and Waller (2022) is an important FEDS note because, in our view, it can be taken as proxy of the Fed staff view. As mentioned, we are agnostic about the slope of the Beveridge curve. In any case, even conditional on the assumptions of Figura and Waller (2022), inflation barely returns toward target in 2024. Indeed, using similar assumptions but different measures of jobs opening and labor turnover produces a core PCE price inflation forecast well above target in 2024. Overall, the conclusion remains the same of the last several months: without help from the supply side, the Fed seems to need either a protracted period of below potential (or negative) growth, or a deep recession. (Indeed, we are not surprised that two quarters of negative growth in 2022 had little impact on inflation dynamics. Models simulations discussed in the last 6-9 months had shown that several quarters of low/negative growth are needed). In a nut shell: the Figura and Waller scenario can happen. But at the moment it seems quite optimistic.

Appendix

Figure 2. Phillips curve-based forecasts

A. Using jobs openings rate

B. Using openings per unemployed

C. Using quits rate

Note: the figure shows the forecasts obtained using our “main” Phillips curve model substituting the unemployment gap with jobs opening rate (panel A), openings per unemployed (panel B), and quits rate (panel C). Shaded areas indicates confidence bands calculated using quasi-out of sample exercises.

Contact

Want something more tailored?

We provide tailored consulting on ad-hoc projects.