The Employment Report: The Fed never “loses”

In his note the day before the employment report, Fed Watcher Tim Duy concluded:

Bottom Line: The Fed would have an easier time paying attention to the weak inflation numbers if the economy was not operating near their estimates of full employment and clearly growing at a pace that will soon surpass those estimates.

Consequently, a report near consensus expectations will tend to strengthen their resolve regarding further rate hikes. A report that falls short of consensus, however, would likely be deemed as noise given the generally solid path of economic activity this year.

We now know that at least some aspects of the report were above consensus, meaning, apparently, that their resolve regarding further rate hikes will strengthen.

Many in the markets echo Fed thinking:

The Fed has to be cognizant of the fact that if we run employment at 4%, inflation could rise above their 2% target,” said Gus Faucher, economist at PNC Financial Services. “If they let unemployment get too low, they may have to raise rates more aggressively, and that risks causing a recession.”

The reasoning is wrong! In fact, the labor market must be much weaker than what is indicated by the official statistics, despite the observation that the unemployment rate has fallen below estimates of the so-called “natural rate”.

Why? In the chart we observe that over the past two years both unemployment and compensation growth are falling. Hard to envision these things being simultaneously true and at the same time indicate the labor market is strong (strengthening or solid). Before that, compensation growth was low and steady while unemployment dropped significantly.

The view that inflation is determined by unemployment, the P-C view, is a “solid” barrier to the advancement of knowledge.

The next charts plots compensation growth and inflation (the PCE core variety) since the time inflation became low and stable. Even when compensation growth trended up, inflation remained on the low and stable track.

Unemployment does not determine inflation. Inflation is a monetary phenomenon, whose trend is determined by monetary policy (the stance of which is given by NGDP growth, not interest rates).

The next chart shows that when NGDP growth tumbled, inflation notched down. With NGDP growth lower than before since the expansion began, inflation is a bit lower than previously (1.9% vs 1.6%).

The next chart shows that in addition to being the main determinant of inflation, NGDP growth is also the main determinant of cyclical unemployment. The “mechanism” here works through the wage/NGDP ratio and wage stickiness.

When NGDP growth falls significantly below trend, or tumbles into negative territory as in 2008-09, the wage/NGDP ratio rises (or shoots up as in 2008-09) and unemployment rises (or soars).

Therefore, “low inflation” goes together with “high unemployment” because monetary policy so determines.

If you look closely at the chart, you will notice that when NGDP growth takes a “breather”, the fall in the unemployment rate slows down (for example, from mid-2015 to early 2017).

Recently, the Fed and others have shown surprise that inflation is not going up towards the target despite falling and “too low” unemployment. That´s just not possible with NGDP growth on a stable and “low path”, although that path is sufficient to get unemployment lower (given the even lower, now falling, rate of compensation growth).

A few years ago, they were surprised that despite the high rate of unemployment, inflation did not fall more.

Fed credibility and anchored inflation expectations explain the “surprising” outcome. Just remember Bernanke´s 2002 speech “Deflation making sure “it” doesn´t happen here”.

With the Fed´s “tightening resolve” intact, the “talk of the town” is yield curve flattening and future inversion. If the Fed goes on as expected, yield curve inversion will happen. The Fed, however, has a standard answer: “This time it´s different”!


Leave a Reply

Your email address will not be published. Required fields are marked *