news article background image
2019-03-29Macro `n Cheese

Running a “High-Pressure” Economy

This week, Roger returns to a subject that he has delved into many a time in this forum, the Phillips-curve. But instead of discussing whether it is dead, dormant or alive and kicking he goes looking for evidence that running a “high-pressure” economy is good policy. How does this relate to the current economic policy debate? – Well, read on!

Back with a vengeance

The Federal Reserve (FED) has made almost a complete U-turn from the December (2018) hike and now seems able and willing to lower the interest rate (and as my favorite FED-watcher, Tim Duy, is arguing here, it probably won’t take much for that to happen). In its latest iteration, the median rate path (yes, the infamous “dot plot”) doesn’t even reach “long-run” (neutral) during the active forecast period, implying that the FED is suddenly fretting an uphill battle on inflation and inflation expectations, especially if the economic outlook would deteriorate below the 2.1%  y/y median forecast for this year:

So, what I see is inflation that's close to two percent but that sort of keeps bumping up against two percent and then maybe moving back down a little bit. And I don't feel that we have kind of convincingly achieved our two percent mandate in a symmetrical way. Now, what do we mean by symmetrical? What we really mean is […] that we would be equally concerned with inflation, persistently above as persistently below the target. So that's really our framework, and I don't think we've quite achieved it yet, because we're really ten years deep in this, almost ten years, in this expansion, and inflation is still kind of, I'd say not, you know, clearly meeting our target.”

(Edited excerpt from the transcript of Chairman Powell’s press conference 20 March 2019)

Today, I thought we could discuss if this clear break from the “Brainard principle” (favoring a conservative – slow-moving – conduct of monetary policy) is good policy or not and what such a rapid change from the world’s leading central bank means for other central banks, should normal cyclical patterns hold.

To begin with, I think it is important to see these developments as, at least partly, warranted by the closeness[1] of the effective lower bound (ELB) and relatedly, as Powell seems to refer to in his statement above, a last-minute push to make the tight labor market count towards raising wage and inflation expectations. If this is the sentiment among US monetary policy makers, who after all have managed to raise rates to the 2¼ - 2½ percent range, it is instructive to consider how policy makers in Europe[2] and elsewhere will react if a new slowdown/recession materializes anytime soon.

[1] In a recession, FED has historically lowered (nominal) interest rates by 4 to 5 percentage points (p.p.).

[2] In a recession, ECB (and its predecessor BuBa) has historically lowered (nominal) interest rates by 3 to 4 p.p., albeit the min-max range appears somewhat wider (2 to 5 p.p.).

Given the perennial discussions on the decline – or possible death – of the Phillips curve in academic and policy circles (not to mention among financial market pundits), I think very few analysts assign a very high possibility of success to Powell and the FOMC’s recurring strategy of running a “high-pressure economy” (cf. Yellen, 2016) to address the acute policy issues stated above.

As previous blog posts should have made clear, I am no big fan of using the Phillips curve as an input to monetary policy analysis; I find the relationship between resource utilization and inflation far too unstable (both in size and time) and far too imprecise to be used for interest rate decisions.

Nota Bene, that does not mean that I don’t think there is any Phillips curve relationship at all[3]. The fact is, right now is probably an excellent time to explore that relationship and lend some support for policy makers willing to (like Powell et al) experiment. What I am thinking of, in particular, is if it is possible that the Phillips curve “re-emerges” whence we have run a sufficiently tight labor market for a sufficiently long period of time?

[3] My long-held belief is that, by and large, when labor input is not scarce – wages are decided by productivity (with an eye on relative measures – competitiveness, especially in the small open economy case. Still pondering the whole “monopsony debate”).

– To test that hypothesis, let’s have a look at what happens with coefficients on wage inflation when we divide labor market developments into three “buckets”: (1) When the unemployment rate is very low (< 1 standard deviation below mean); (2) When the unemployment rate is very high (> 1 standard deviation above mean), and finally; (3) When the unemployment rate is in a “normal” range (+/- standard deviation around the mean). Essentially, what we are doing is estimating segments of the Phillips curve.

 

Note: These divisions can be made in a number of more intricate ways using absolute numbers (NAIRU), unemployment gaps, CPI inflation or more subdivisions etc.., which is why you should of course feel free to download, use and tailor the below Macrobond chart/analysis to your own preferences. But first, let’s look at a broader timeframe, going back some 30 years.

Unsurprisingly, in a longer time frame[4], the relationship seems to be well functioning, quite linear and the coefficients are stable (a tad low maybe?), except for perhaps at periods with very high unemployment rates. In general, each percentage point of a lowered unemployment rate increases AHE-growth by around 0.35 percentage points.

[4] I have used a during/after the great moderation sample to keep inflation expectations somewhat stable. But if you pretend to look at even longer time frames, I think it is best to include some control for inflation expectations as it has been proven an important factor in many Phillips curve studies.

But looking at a shorter time frame, let’s say the past 20 years, something interesting happens:

 

The statistical relationships strengthen at lower unemployment rates (and vice versa at all other time frames), and at really low unemployment rates, the inflationary impact seems to become gradually more pronounced. We could of course look at shorter time frames still, with many studies suggesting a break around 2004 (or post-crisis if you prefer), but to retain some aura of statistical seriousness I must underline that we simply have too few observations in our calculations, making the intercepts jump around, etc. etc. Now, we would prefer to complement the analysis with some cross-sectional estimation method[5] (applied on regional data, e.g., like in this) and/or even some non-linear estimation method. But just to give you an idea:

[5] From what I hear, it is in the workings.

 

So, what does this all tell us? – Well, one could simply state that the statistical Phillips relationship is becoming more convex, and several recent studies are of course utilizing non-linear methods instead. However, such methods, by construction, need to step away from the coveted simplicity of the standard Phillips curve framework and add additional theoretical layers (capacity constraints, price adjustment asymmetries etc.). Not to mention that I have yet to find a robust non-linear specification of the Phillips relationship. And, either way, even non-linear models do little to change the impression that the Phillips relationship is moot, or does not hold, for rather long periods of time. (…And all of a sudden I came to think of this.)

Nonetheless, the results are also conducive to trying to run a high-pressure economy to overcome current monetary policy challenges: The beta coefficient for the below-mean unemployment rates seems to have strengthened considerably over recent years. This suggests that even though the Phillips relationship might have become more complex over time, managing to push unemployment rates down from more normal levels (by whatever means necessary – yes, I’m looking at you, fiscal policy!) could simultaneously provide a fast and simple way for policy makers to increase the distance to the effective lower bound. In short, Powell et al. are probably on to something when thinking about running a high-pressure economy!

March of the lemmings

So far, we have only discussed the US situation, but from most aspects the situation is even more concerning in other important economies. Recently the ECB has come under criticism for being over-active in announcing new TLTROs and tiered deposit rates etc. Let’s apply the same framework on them and in the process, maybe, we have created an analytical tool flexible enough for you to apply on your favorite central bank[6]!

[6] I cannot help but smile (only on the inside) when pondering the a similar context for a central bank near me. Always getting itself in trouble.

Data available for the Euro area has not been completely aligned to ECB and market expectations, but rather on the weak side. Despite a noticeable (but still too cautious) relaxation of the fiscal stance in almost all Euro-countries. Is there any way in which the ECB could employ the same strategy the FED is using? – Well, looking at the full period (with established inflation expectations, ca. 1992) and taking into account the longer lags (due to, i.a., stronger rigidities) involved there does indeed seem to be a way forward:

 

In comparison to the US, the full sample coefficient is very high indeed and suggest that wage pressures might still do the trick for ECB. That said, narrowing the sample to a more recent time period (from ca. the introduction of the EUR), indicates a less explosive but still promising way to use a monetary policy “pressure cooker” in the Euro Area as well. (Again, note the problem with small samples when dissecting the Macrobond-file).

 

I’ll leave you with a couple of final thoughts:

  1. While we can all take argument with what exact stabilization policy measure should be employed each time, I think the FED is right to “over-react” at the first signs of weakness – the distance to the ELB is and should be a factor in how a central bank reacts, especially in the current environment.
  2. While the Phillips relationship might not be a precise instrument for monetary policy calibration, its prevalence at high levels of resource utilization offers one of our best hopes of putting some distance between ourselves and the effective lower bound. – Central banks are wise to jump on the bandwagon.

 

 

Get these posts delivered to your inbox

Sign up now to get an email every Friday with the latest research posts from Macro n Cheese

Sign up

 

Disclaimer

We don’t usually have views and opinions about economic and financial states of affairs, (not ones that we express publicly as a company, anyway). We do believe, however, that people can and do appreciate a variety of perspectives. What you’ve just read is the perspective of our resident chief economist. While we think he’s very smart, Macrobond Financial does not expressly endorse the views he presents here. And, as the old adage goes, you shouldn’t believe everything you read (not without finding the data, performing a few analyses and presenting it in a nice chart). We want to make it clear that we are not offering this information as investment advice. That being said, if you have the application you can easily check everything that’s mentioned here, and decide for yourself. If you don’t have the application, now you have a great reason to get it.