Will flagging productivity growth trigger a hawkish response from the Fed? That is a question I have been asking myself since Federal Reserve Chair Janet Yellen discounted the cyclical influences of low wage growth in her July 10 speech:
The most important factor determining continued advances in living standards is productivity growth, defined as the rate of increase in how much a worker can produce in an hour of work. Over time, sustained increases in productivity are necessary to support rising household incomes…Here the recent data have been disappointing. The growth rate of output per hour worked in the business sector has averaged about 1‑1/4 percent per year since the recession began in late 2007 and has been essentially flat over the past year. In contrast, annual productivity gains averaged 2-3/4 percent over the decade preceding the Great Recession. I mentioned earlier the sluggish pace of wage gains in recent years, and while I do think that this is evidence of some persisting labor market slack, it also may reflect, at least in part, fairly weak productivity growth.
Goldman Sachs economists led by Jan Hatzius hypothesize that productivity growth is low only because growth is mis-measured, undercounting the value of free or improved software and digital content made possible by information technology (although see Greg Ip’s opposing view). It seems that Yellen is leaning in the direction of taking the productivity numbers at face value and seeing low wage growth as consistent with the view that the productivity slowdown is real.
Indeed, the productivity trends may be even grimmer than 1-1/4 percent that Yellen cited in her speech. Consider for example two measures of underlying productivity growth trends, a moving average measure and the result from a simple unobserved components model:
The downtrend since 2000 is more easily discerned by focusing only on the trends:
The unobserved component approach suggests that productivity growth decelerated to an annualized pace of just 0.82 percent by the second quarter of this year. That is interesting because it must be somewhat similar to the Fed staff estimates. The accidentally released Fed staff estimates of potential GDP growth range from roughly 1.6 to 1.8 percent through 2020. That is exactly what you might expect with productivity growth of around 0.8 percent and labor force growth in the between 0.8 to 1.0 percent (roughly the recent range). It seems the Fed staff have adopted the pessimistic view of the productivity numbers. It is not about measurement error.
What are the implications for policy? Yellen might think back to the 1990’s, when a surprise rise in productivity growth temporarily lowered the natural rate of unemployment. Slow wage adjustment meant higher than expected productivity growth had a disinflationary impact on the economy, allowing for then Federal Reserve Chairman Alan Greenspan to hold interest rates relatively low. Yellen might reverse that logic now and think that the arguments for tighter policy are stronger. Hence another reason why low wage growth is not an impediment to raising rates.
Any increase in the natural rate of unemployment, however, would be temporary. After wage growth adjusts, we would expect the equilibrium real interest rate to fall in response to lower productivity growth. Lower productivity growth reduces the expected return on capital, thus requiring lower interest rates to maintain neutral policy.
The trick then is determining whether we are still in the short-run or already in the long-run. It may be that wages have already adjusted. Real wage growth was fairly high during the recession and it’s aftermath:
Real wage growth – using core-PCE as the deflator – was high as wages adjusted during the recession, but then remained high even when unemployment was above six percent. Now real wages are running low. Could it be that the wage adjustment to lower productivity growth occurred during the recession? If so, the Fed would be in error to believe that now is the time to tighten policy in response to low productivity growth. In effect, the zero bound unintentionally did that job for them. Not only did wage growth adjust, but any potential inflationary impacts were overwhelmed by the disinflationary impacts of the recession.
The bond market, with ten-year Treasury yields hovering between 2 and 2.5 percent, appears to be fiercely discounting the lower-for-longer story consistent with low productivity growth. Furthermore, low TIPS-based inflation expectations and a modest expected path for short rates also suggest little need for the Fed to tighten policy to avoid a 1970’s style inflation. The FOMC, however, has a more hawkish view, anticipating a more aggressive policy over the next few years. The Fed staff split the difference in their forecasts. The direction the FOMC ultimately takes could be the difference in an expansion that last four more years or only two.
Bottom Line: Fed policy increasingly reflects the view that the productivity growth slowdown is real. We see it in falling estimates of potential GDP growth, falling expectations for the terminal federal funds rate, and now we see it as a reason to anticipate low wage growth. The first and third reactions seem to have had a hawkish impact on policy - not only is low wage growth not an impediment to raising rates, but San Francisco Federal Reserve President John Williams argued the Fed needs to engineer a substantial slowdown in growth next year. But the FOMC has yet to act on that relative hawkishness; to date they have moved in the direction of market participants. Indeed, while I suspect the odds favor a September hike, we don’t even know they will raise rates this year at all! The question is whether they would be quick to act on that hawkishness in the face of any unexpectedly high inflation or wage growth numbers. I am thinking low-productivity growth coupled with memories of the 1970s may prime FOMC members in that direction.
