February Payrolls: Nobody Panic
Josh Wright, Chief Economist | Economist Corner
Friday, Mar 08, 2019

If the Fed needed any more excuses for another dovish meeting, that need has now officially been filled. The response to the February jobs report from most commentators has been refreshingly grounded – calls of “Don’t panic” abounded – but a payroll print of just 20k thickens the mud in recent economic data more than enough. The decline in the unemployment rate back down to 3.8% (from 4.0%) and a bounce back in hourly earnings growth (3.4% on the year, 0.3% on the month) were there to reassure us all the sky wasn’t falling, but there were other sources of comfort in the details.

Here are a few highlights:

  • Focus on the trend, cancel out the noise: this report reminds us that trend job growth should be softening as fiscal stimulus fades… but not this much. The 3-month moving average landed right where economists expected the 1-month number to be: around 185k. Also, the very big misses in monthly job growth over the last few years have usually been revised upward. No fun to point out, but true.
  • The big grains of salt: Could this be a sign of something more drastic, a real turn that won’t get revised up? Maybe, but then why were payrolls so strong in January, when other economic data were already stumbling? There are too many signs of distortions to see this month as a crucial moment of convergence. The more likely case is that the labor market continues to trundle along, as it did in early 2016 (not without the occasional stumble).
    • Shutdown-related distortions: part-time unemployment and the U6 “under-employment” rate plunged after spiking in January as furloughed workers apparently took on temporary employment to keep cash coming in. The spread between U6 and the U3 unemployment rate has returned to its post-recession low (3.5%), although it still has another half a point to go before it reaches the lows of late 2000 (2.9%). That suggests there may yet be room for improvement (i.e., pockets of slack), despite mixed messages on labor participation this month.
    • Weather-related distortions too: construction (-31k vs 53k prior) and leisure/hospitality (0k vs. 89k) were particularly weak. Construction may be particularly susceptible to weather-related swings at this time of year. Less construction in winter --> smaller sample --> more noise in sample data and the seasonal adjustments.
    • Despite the weak headline payroll number, the diffusion index of 57.2 looks decent, although well below the preceding 60+ prints.
  • The biggest concerns in the payroll survey:
    • The decline in weekly hours, which undermines the rise in hourly earnings (less total take-home pay).
    • The drop-off in the breadth of manufacturing gains may be a bellwether, although it was encouraging to see the headline for manufacturing remained positive (4k vs. 21k).
  • The market is right: this report is not the stagflationish nightmare for the Fed that it might seem. The Fed has already put itself on hold for the next few months. They won’t be quick to change their stance unless a clear signal emerges from either financial markets or the economic data. This is anything but, and it’s an easy call to wait for the dust to settle on Q1 before making any significant moves. Even the wage growth will be easy to talk down, given the decline in hours. That said, this report does raise the stakes on inflation data and inflation expectations. And slower growth should bring a higher frequency of questionable employment reports in the months ahead.

about the author

As chief economist at iCIMS, an industry-leading provider of talent acquisition software, Josh Wright leads a team of data scientists in analyzing emerging trends in the U.S. labor market