Jack Dunn - Reclaiming Common Sense

January Jobs: The Silent Revisions, Plural.


The January Employment Situation Report, or Jobs Report, was released this past Friday, one week ago. There was a considerable amount of data to digest. The Current Employment Statistics (CES) worker data saw considerable revisions, mostly to the data from 2015 through 2017. These revisions created a surprise. The Current Population Survey data was unchanged for the same period of time. The surprise here was the dramatic upward revision to the workforce population value.  This upward revision negatively impacted the workforce participation number. So far four articles have been written:

  • "January Jobs Surprise" details how the Seasonally Adjusted (SA) Private Sector Workers data could have been reported up 452,000 workers, instead of the 196,000 that was reported.
  • "Seriously Strong CPS Data" detailed how we had one of the best January reports since 2003 with minor drops in non-seasonally adjusted full-time and part-time jobs and a minor uptick in unemployment. This is normal to see drops in jobs and drops in workers, non-seasonally adjusted, during January. It is one of the reason that the data is seasonally adjusted.
  • "Five Presidents after 12 months: Jobs" compared President Trumps data to Presidents Reagan, Clinton, George W. Bush, and Obama. President Trump and President Clinton had great first years - each adding approximately 2.3 million jobs. President Clinton added 2.3 million net part-time jobs and President Trump added a net 2.3 million full-time jobs.
  • "January Jobs up in 10 of 11 Sectors" details how all sectors lost non-seasonally adjusted workers, ten of eleven sectors gained workers, and how the data from 2015 was revised lower, while the data from 2016 and 2017
  • "January Jobs: Battle of the Sexes" could have been titled "Men Not Working." Fewer men were working full-time jobs during January 2018 than July 2007. That is not a typo. Ten plus years later fewer men are working full-time jobs.

Add to the data that was massively revised this past month was the multiple jobholder data from 2012-2017, except for December 2017. Why does this matter? Multiple job workers "depress" participation by having some people "over-participate." Multiple job workers do not receive "less-employed" benefits when they lose their jobs. 


We had a near record level of Multiple Job Workers for the Month of January. You did not hear this in the mainstream media because "all" they care about is the headline number of "seasonally adjusted jobs" created and the seasonally adjusted unemployment rate. What they are really reporting is the number of workers added to the economy. January is a job loss month, non-seasonally adjusted. January is a worker loss month, non-seasonally adjusted. If there were no seasonally adjusted data there would have been no "jobs streak" under any President.  The number of Multiple jobholders has depressed participation by having some people over-participate and some people under-participate. We had over 7.75 million people work multiple jobs this past month. 


More People Working Two part-time Jobs.  There are four basic ways that a person can work multiple jobs, only three of them have their own data categories. People can work two full-time jobs. People can work two part-time jobs. People can work a primary full-time job and a secondary part-time job. The keepers of the data also measure a "hours vary by period" category. When all of these categories are combined it does not equal the total multiple jobs worked category.This column has reported a primary part-time secondary full-time value as the remainder from the Total MJH minus the FT-FT minus the FT-PT minus the PT-PT worker data.  We had over 2 million people working two part-time jobs during January of 2018. The first time we ever recorded over two million people working two part-time jobs was during April of 2009. We saw two months of PT-Pt over two million during 2012. That happened three times during 2013, four times during 2014, three times during 2015, eight times during 2016,  and five times during 2017. This column has hypothesized that this is an unintentional consequence of the Affordable Care Act, also known as Obamacare.


We had a January Record number of people working two full-time jobs. There were 339,000 people working two full-time jobs last month. That number during 2017 was 302,000. We had only eclipsed 300,000 FT-FT workers once before that time and that was during 1999.


The seasonally adjusted total multiple job workers data was revised from 2012 through November 2017. Note that the underlying non-seasonally adjusted total multiple jobholder data were not revised. This is part of the reason why this column focuses on the non-seasonally adjusted data. The seasonal factors used to convert non-seasonally adjusted data from month to month, season to season, and year to year. What changed regarding the data between 2012 and last month that needed massive revisions to the seasonally adjusted data without touching the non-seasonally adjusted total multiple job holder data?


We saw massive revisions to the Current Employment Statistics data last month. This data, the multiple job holder data, is found in Table A-16. This is part of the Household data set. This is where the Current Population Survey data is found. The B-Tables are the employer data and are derived from the Current Employment Statistics data set. It was thought that the only datum that was changed in the CPS data was the workforce population growth rate. This impacted the workforce participation rate data. It turns out that the population datum was not the only data revised in the CPS data set. The most notable changes were January through June, plus August, of 2017. It is also noticeable that he data from January through May were predominantly revised lower and the September through December data were revised higher than originally reported during the December Jobs Report.


There is much more to the Employment Situation report than just the headline "jobs" number and the headline unemployment data. We had one of the best January Jobs reports released on the same day as the "Nunes memo." that is unfortunate. It is unfortunate because the data was remarkable. We had one of the best NSA CES private sector worker growth rates that we have seen since 1981, in that we did not contract as much as we could have contracted.We did not see jobs decline as much as they normally do contract during January, as they "always" contract. We saw the best January for Men and for Women for full-time jobs since 2007. There is general sector growth. We have people working multiple jobs.


It's the economy.