Unemployment, Wages, Inflation, and Fed Policy, by Tim Duy: I apologize if that was a misleading title. This post is not a grand, unifying theory of macroeconomics. It is instead a quick take on two posts floating around today. The first is Paul Krugman's admonishment to the Federal Reserve against raising interest rates before wages rise:
So far, no clear sign that wage growth is accelerating. Even more important, however, wages are growing much more slowly now than they were before the crisis. There is no argument I can think of for not wanting wage growth to get at least back to pre-crisis levels before tightening. In fact, given that we’ve now seen just how dangerous the “lowflation” trap is, we should be aiming for a significantly higher underlying rate of growth in wages and prices than we previously thought appropriate.
I don't think that you should be surprised if the Federal Reserve starts raising rates well before wage growth returns to pre-crisis rates. I think you should be very surprised if the Fed were to do as Krugman suggests. Historically, the Fed tightens before wages growth accelerates much beyond 2%:
As I have noted earlier, wage growth tends to accelerate as unemployment approaches 6 percent, and so if you wanted to be ahead of inflation, they would be thinking about the first rate hike in the 6.0-6.5% range. That 6.5% threshold was not pulled out of thin air.
The second point is that the tightening cycle is usually topping out when wage growth is in the 4.0-4.5% range. One interpretation is that the Fed continues to tighten policy to prevent workers from gaining too much of an upper-hand, thereby contributing to growing wage inequality. Of course, I doubt they see it that way. They see it as tightening monetary conditions to hold inflation in check. Either way, the end is the same. It would represent a very significant departure from past policy if the Fed waited until wage growth was at pre-recession rates before they tightened policy or if they allow conditions to remains sufficiently loose for wage growth to eventually rise above pre-recession rates.
If you want the Fed to make such a departure, start laying the groundwork soon. The best I can offer is my expectation that Fed Chair Janet Yellen is more inclined than the average policymaker to wait until wages actually rise before acting. I have trouble believing that even she would wait until wage growth accelerates to pre-recession trends.
Second, the Washington Post's Ylan Mui has this:
But a funny thing happens once unemployment hits 6.5 percent: The behavior of inflation starts to become random, as illustrated in this chart by HSBC chief U.S. economist Kevin Logan.
The black line represents the average annual unemployment rate for the past 30 years. You can see that in all but two cases (both of which were temporary shocks), inflation declined when the jobless rate was above 6.5 percent. But when unemployment rate fell below that point, inflation was almost as likely to increase as it was to decrease. In other words, what happens to inflation below the Fed’s threshold is anybody’s guess.
I would take issue with the idea that inflation behavior becomes "random" at unemployment rates below 6.5%. You need to consider this kind of chart in the context of expected inflation and expected policy. If inflation expectations are stable, and if the Federal Reserve provides policy to ensure that stability, you would expect random errors around expected inflation. Couple this with downward nominal wage rigidities, and you should expect the same even under circumstances of high unemployment. Here is my version of the same chart:
The data is monthly. This y-axis is the change in inflation from a year ago, where inflation is measured as the year-over-year change in core-pce. Unsurprisingly, since 2000, changes in core-inflation vary around zero. Stable and low inflation expectations. During periods of the 1970's and 1980's you see the impact of unstable expectations as the relationship circles all over the place. But you also see the general pattern of disinflation since the early 1980's with the downward sloping relationship and many inflation observations, even at low unemployment rates, below zero.
Now it is fairly easy to put both of these posts together. The Fed, wanting to ensure stable inflation expectations, begins raising interest rates well before wage rates begin rising. This is turn controls the growth of actual inflation so that inflation rates do not rise as unemployment falls further. The deviations of inflation from expectations are then just noise. But actual inflation is not "random." It is the result of specific monetary policy.
Bottom Line: If the Fed follows historical behavior, they will beginning tightening before wages rise and in an environment of low inflation such that inflation remains stable even as unemployment falls. In other words, in recent history that have not exhibited a tendency to overshoot. Explicit overshooting would represent a very significant shift in the Fed's modus operandi.