Why is U.S. Productivity Growth So Weak?

Former Global Chief Economist for JPMorgan Chase & Ph.D in Economics

Why is U.S. Productivity Growth So Weak?

One may need to look no further than the trend in U.S. employment growth to answer this question.

Non-Farm productivity growth surged by 10.3% during Q2:2020 as U.S. employment growth plunged as the U.S. economy shut down to battle the Covid-19 pandemic.  With fewer labor inputs, the economy enjoyed a temporary boost in productivity growth.

Fast forward to 2022, employers went on a hiring spree to meet strong demand supported by $10 trillion of fiscal and monetary stimulus injected into the U.S. economy during the pandemic. 

Unfortunately, firms tried to hire even more workers than they could find while output remained constrained by shortages of raw materials and other related supply-chain disruptions.  These issues prevented companies from using labor inputs efficiently.  The U.S. economy also suffered sluggish growth during H1:2022 and resulted in a weak -4.1% (after Sept 1,2022 revision) in Q2:2022, SAAR non-farm business productivity reading.  

What Happens if Job Growth Was Even Stronger than Initially Assumed?

When that happens during an episode of sluggish growth, productivity growth will deliver a downward surprise which is what happened during H1:2022.   The Bureau of Labor Statistics recently reported that job growth will be revised higher according to its preliminary estimates, namely +39k per month for the year ending in March 2022!  The BLS will release its final employment revisions in February 2023.  

Some observers still believe that the underlying trend is much weaker because employment levels in the Household Survey (HS) have softened relative to the Establishment Survey (ES).   From March 2022 to July 2022, the ES created +1.7 million jobs while the HS lost -168k jobs for a gap of 1.9 million jobs.   However, we believe that these labor market observers are being creative in selecting March 2022 as their basis for comparison because if they chose December 2021, the gap would have been a more moderate 1.2 million jobs. 

Comparing the employment flows of the HS and ES Surveys is challenging given the different ways each survey measures employment.

What are the Major Differences Between both Employment Surveys?

•        Agricultural workers are included in the HS and not in the ES.

•        Workers with multiple jobs are counted multiple times in the ES but only once in the HS.

•        Self Employed Workers are included in the HS and not in the ES.

•        Unpaid family workers are included in the HS and not in the ES.

•        Private Household Workers are included in the HS but not in the ES.

What if Both Employment Surveys Were Computed Identically?

Interestingly, the U.S. Labor Department computes an HS employment measure using the same assumptions used by monthly non-farm payroll estimates for the ES.  This methodology places the employment metrics on a level playing field.  Now, the “adjusted HS” is up (+3.5 million YTD) and exceeds the ES measure (+3.3 million YTD)!  Using this rigorous method one can no longer say the HS employment metric is weaker on a YTD basis.  Even on a year-over-year percentage basis, the adjusted HS is up +4.5% versus a rise of +4.2% for the ES!

Source: The Bureau of Labor Statistics

What if we Could Measure the Entire U.S. Labor Market?

We know that the Bureau of Labor Statistics (BLS) measures only a subsample of the U.S. labor market to generate its employment estimates.  Bright Query (BQ), Inc. has data on 41 million U.S. Private companies and 4,000 U.S. Public companies.  Since the data is not seasonally adjusted, we analyze all employment figures on a year-over-year percentage basis to eliminate any issues with seasonality.  BQ’s data begins in January 2010 through the present (i.e., July 2022).  BQ also has revenues, profits, and net income data for all these companies. 

Without the sample size constraints, we believe BQ’s employment data is more accurate than the other monthly employment metrics generated from the HS-adjusted and ES sources.   Interestingly, BQ’s data reveals +5.6 million jobs created on a YTD basis compared to 3.5 million for the adjusted HS and 3.3 million for the ES, respectively.   

BQ’s faster employment growth in 2022 sheds some light on why the market consensus underestimated the growth rate in non-farm productivity growth in H1:2022. 

Source: The Bureau of Labor Statistics and BQ

Source: The Bureau of Labor Statistics and BQ

Unfortunately, the massive hiring during H1:2022 was unable to produce a corresponding increase in output due to supply-chain disruptions and demand destruction due to higher prices.  In housing, homebuilders reported shortages of lumber, garage doors, windows, and other materials needed to complete the construction of new homes previously sold.  Next, the housing sector suffered some demand destruction due to affordability issues resulting from higher mortgage rates and rising prices.  In the auto sector, car production tanked due to semi-conductor shortages needed to complete the assembly of many pre-ordered vehicles.  Weaker output led to negative real GDP growth, and the increase in labor inputs generated a stronger than expected contraction in U.S productivity growth. 

What Are the Advantages of Using BQ Data?

Without sample size constraints, BQ’s data can pick up labor market inflection points. We observed that many U.S. workers left their jobs after the Covid-19 pandemic to start small businesses or join smaller start-ups.  This is not captured by the ES!  The only mystery in the data is why the experimental HS-adjusted series remains so strong.  One reason might be due to errors measuring the labor market incurred with a limited sample. 

Will U.S. Employment Growth Accelerate or Decelerate Within the Next Year?

With a contractionary monetary policy in place, as reiterated by Fed Chair Powell, and the slowing of M2 growth along with quantitative tightening (QT) poised to accelerate in September 2022, we expect all employment measures will slow within the next 12 months.

The U.S. M2 money supply (see Chart below) has fallen from a peak yearly growth rate of 27.5% on Feb. 22, 2021, to a growth rate of 4.8% as of August 22, 2022!  This deceleration will slow economic and employment growth as the Fed continues to raise interest rates to battle inflation.  Conventional wisdom assumes that monetary policy impacts the economy with a lag of 6 to 18 months.  Since the first 25 basis point Federal Reserve policy rate hike occurred on March 17, 2022, we should expect to see the initial effects of this action beginning in Q4:2022 and beyond as the Fed has continued to raise interest rates.   The good news is that we have already seen some preliminary evidence of a peak in inflation (see our previous newsletter discussion on this topic).