The Strange April Jobs Report

Before getting into April’s Jobs Report released by the Bureau of Labor Statistics today, this piece by Neil Irwin of the New York Times explains exactly why each initial release of the Jobs Report should be taken with a grain of salt.  In short, attempting to measure the employment picture in a country with 300+ million people is difficult to do and subject to sampling error that can produce wildly different conclusions.  Reading too far into one jobs report is foolhardy.  The timing of Irwin’s article turned out near perfect as April’s report highlighted this point well.

Here were economist expectations before the release of the Jobs Report:

• Payrolls: +218k
• Unemployment: Tick down a tenth to 6.6%

And here are the reported numbers:

Headline Numbers:

• Payrolls: +288k
• Unemployment: 6.3%

Other Important Numbers:

• Net Revisions from previous months: +36k
  • Feb: From +197k to +222k
  • March: From +192k to 203k
• U-6 Broad Measure of Unemployment:
  
  • Down from 12.7% to 12.3%
• Labor Force Participation Rate:
  
  • Down from 63.2% to 62.8%
• Household Survey:
  
  • Employment: -73k
• Average Hourly Earnings Flat
  
  • $24.31
• Average Workweek Flat:
  
  • 34.5 hours

Our Take:

On first glance, headline numbers look great and it would be easy to conclude April was a great month for the economy – payrolls exceeded expecations with a robust 288k gain, revisions to the past two months netted 36k more jobs than previously reported, unemployment fell drastically to 6.3%.  Upon digging deeper, however, things get bizarre.  The household survey showed that the economy subtracted 73k jobs last month and the participation rate fell 4 tenths.  The market appeared to react to the report in this order: Great headline numbers! Oh, the internals…

The jobs # was good for stocks until, you know, it wasn’t. pic.twitter.com/Qr45WGoiwR

— Nicholas Johnston (@FirstWordNick) May 2, 2014

It is hard to reconcile how the economy added 288k jobs last month yet the number or people working fell by 73k.  This is the time to heed Irwin’s advice.  The payrolls number comes from the establishment survey that polls businesses on their hiring practices.  The household survey polls households on whether they were able to find work in the previous month.  Most times when these surveys conflict, it is simply due to sampling error and the picture becomes clearer in the following months as revisions to initial data are made.

We always warn readers to beware the noise in monthly numbers. But this month more than usual. http://t.co/MzwkzuWljI

— Ben Casselman (@bencasselman) May 2, 2014

It is better to look at long term trends as they paint a clearer picture of the direction of the economy and the trend for some time has been slow, consistent, boring growth.  Anemic winter jobs growth and the lack of GDP growth in the first quarter caused some to worry whether this trend was worsening.  This report puts those concerns to rest for two reasons: 1) 288k is a fairly strong number of jobs added and 2) revisions to February and March data show that we now have seen 3 straight months of 200k or greater growth in the labor market.  There is certainly no takeoff yet, but a clear upward trend remains.

The year-over-year trend, updated with today’s data. Still no clear acceleration, but looking better at the end. pic.twitter.com/mEH7YIo0a3

— Ben Casselman (@bencasselman) May 2, 2014

While this report is bizarre and mixed, it is positive on net and supports the storyline that the economic recovery will pick up steam in the second half of 2014.  The Fed appears to have made the right decision by staying the course and continuing the taper in May.