Jack Dunn - Reclaiming Common Sense

Seasonal Factors and Upward Revisions to May and June Data Mute July Gains


Multiple job workers could have increased or decreased by 2%The Monthly Jobs Report can be a market mover.  The jobs report is created using two surveys, the current population (CPS) survey or Household Survey, and the current employment statistics (CES) survey or Establishment Survey.. The CPS data measures the number of full-time jobs, the number of part-time jobs, and the number of unemployed workers, as well as the workforce population. The CES data collects information on the number of workers in the workforce as well as wage data. The headline number is the seasonally adjusted non-farm payroll (NFP) worker data. The number that has been used for the Job Creation Streak is the private sector worker number. The difference is that the NFP data includes government jobs.  July is not a good month for government jobs and was projected to possibly decrease the headline number. Every data set has a seasonally adjusted (SA) and a non-seasonally adjusted (NSA) component. The seasonal factors used to convert the NSA data to the SA data change by category, month, and year.


The abundance of data allows for multiple articles. The July Jobs Forecast article projected that:

  • Full-time jobs would spike
  • Part-time jobs could decline
  • Unemployment would most likely increase non-seasonally adjusted, and possibly decrease seasonally adjusted, as more people look for work
  • The non-seasonally adjusted workforce participation rate should exceed that of July 2014, 2015, 2016 and 2017.
  • The month to month NSA CES private sector growth indicated that we could see 240,000 to 330,000 workers added. A similar range was indicated by the rolling year growth rate. It was thought that the more appropriate range was under 295,000 and more likely to be 265,000.
  • There should have been non-seasonally adjusted worker growth reported in all sectors except government, and education/health services. Information technology should have shown annual growth even with a month to month decline.
  • Weekly Wages should continue to rise as higher paying jobs expand their workforce and as more full-time jobs are created
  • Upward revisions to the May and June data would "borrow" growth from July..

What was recorded (NSA) and what was reported (SA?)


The headline seasonally adjusted worker number was up "only" 157,000 workers. Here's the thing: the June SA CES NFP was revised up by 47,000. If we add the bounce to the July data then we see a value of 204,000. This is still a little lower than what was expected. The Seasonal factor was also low.  The total private sector number was 170,000. The SA CES Private sector data was revised up 55,000 workers. This means that number could have been reported at 225,000. We could add another 6,000 if we used last year's seasonal factor. A number of 231,000 is more realistic. The reason for the difference between the NFP number and the Private sector number is because Government jobs dropped significantly. The NSA NFP recorded a drop of 1.156 million total workers. Those are the main takeaways, right now, for the CES Data. The worker trends will be addressed in the Wages and Workers article.


Unemployment Dropped Month to Month and July to July.  This Month's 6.73 million was a drop of 82,000 non-seasonally adjusted unemployed workers. We have fewer unemployed workers than any other July back to July 2000 when there were 6.028 million. This NSA U-3 is comparable to the July 1989 level of 6.736 million. Remember, we have 257.8 million in the workforce population and million 163.7 million participants. July 1989 we had 186 million in the workforce population and 124 million participants. We have 70 million more workers, 62 million more participants, and 6 thousand fewer unemployed workers. 


Full-time Jobs jumped by 707,000 while part-time jobs were trimmed by 169,000. This is not record setting and it is not a sharp stick in the eye. There were more jobs added this July than July 2009, 2010, 2011, 2012 and 2013. It was also better than July 2017.


Participation Matters. We have the highest July participation rate since July 2014 when we were at 63.53%. The trend is our friend. The 2016 peak participation was up July 2015 to July 2016, July 2016 to July 2017, and July 2017 to July 2018.


Workforce Population Matters. The workforce population grew at an annualized rate of 1.06%. Participation grew at 1.12%. During 2017 the growth rates were 0.60% for population and 0.75 for participation. The 2016 values were 1.09% and 1.37%. During 2014 and 2015 the population was growing faster than the participation. The 2014 values were 0.92% population 0.24% participation and 1.15% and 0.61% during 2014. The population data is "arbitrary." It impacts the participation rate. If the population was growing at the same rate as last year then the participation rate would be increasing faster than it is being reported. Remember that there was a major revision to the population level down January of 2017 and up January 2018.


Multiple job holders increased during July. The multiple job holder data is the CPS data. The participation rate is calculated by the total number of PT and FT jobs, not workers. The NSA multiple job holder level popped up from 7.457 million to 7.833 million. The seasonally adjusted value rose from 7.619 million to 8.072 million. 


This column will produce more articles based on the Jobs Report Data. The wage data has been undervalued in the seasonally adjusted data. The sector data will also be analyzed with the wage data in the "Wage and Worker" article. How is President Trump doing compared to his predecessors? That will be covered in the "Five Presidents at 18 months" article.  This column will also address the changes in jobs and unemployment for men an women, analyze the trends in our aging workforce, and examine the multiple job worker data in more detail in individual articles.


This was a solid jobs report. The uptick in private sector jobs was impacted by a significant drop in government jobs. This will resolve itself later this year when there is a surge in government workers. We are on track for one of our best years ever for NSA CES private sector worker growth. Unemployment, Continuing Claims and U-3 Unemployment, tend to bottom during October. Will we break below 6 million again? This wasn't a record setting month. This was another solid month.


It's the economy.