October Should be a Good Month for Hiring
The monthly employment situation report for September was an interesting report. The "official" seasonally adjusted (SA) non-farm payroll (NFP) Current Employment Statistics (CES) number fell. The reality was that the non-seasonally adjusted (NSA) private sector number surged. We also saw a spike in the non-seasonally adjusted (NSA) Current Population Survey (CPS) jobs. We also saw a drop in the unemployment level, an increase in the participation rate, and a drop in the number of people working multiple jobs. This column wrote a series of articles on the September Jobs Report. The article "Our Aging Workforce this September" has links to all of the articles written including the Sector article, the Multiple Job Holder article, and "War on (Wo)Men" article and the "Five Presidents" article. There was another article written on the September Jobs Report as it relates to the hurricanes and the jobs and unemployment situation in Texas, Florida, and Puerto Rico.
October is a Good Month for Workers. We have seen the October worker level grow by 0.38% to 0.56% from September to October for the past seven years. We have seen a considerable amount of hiring people who already have jobs. We could see another spike in multiple job holders through December. Last month there was a line of discussion that the CPS data, known as the Household data, found people at home who were finding jobs while the CES data, also known as the Establishment Data, that showed that the Employers paid less workers than they paid during September. It is very possible that the downturn in the NSA CES data will be revised higher, without fanfare.
October is a Crapshoot for the Seasonal Factors. The NSA CES data is recorded at one level and multiplied by a seasonal factor to obtain the SA CES value. The seasonally adjusted data attempts to smooth out the peak months for hiring and fill the valleys of job cutting. It is this variation in the seasonal factors that makes comparing seasonally adjusted data dubious.
Cut out the Combination Table and use it as a dartboard. If we grow like we did during 2013 and use the same seasonal factor then we could add 139,000 workers. If we grow like we did during 2014 and use the seasonal factors from 2014 then we could add 155,000. If we use the 2016 seasonal factor for the same growth rates we could add 158,000 to 244,000 workers. Once you have picked the growth rate and the seasonal factor remember that there will be adjustments to last month's data. Is it the situation of a net addition of 7 seasonally government workers or a net gain of a seasonally adjusted 1.3 million government workers. How did we lose lose thousands of seasonally adjusted workers and add a million part-time jobs? How is down up or up down? What is going to happen after the revisions to last month's data are processed? If last month's data is revised up by 100,000 workers that will "steal" 100,000 workers from October. Will the NSA CES Private Sector data return to its path following President Clinton's worker growth pattern?
October should record the addition of hundreds of thousands of Jobs. Hiring normally ramps up for the Christmas Shopping Season. Notice from the NSA CPS histogram that we "always" add part-time jobs. There were not many PT jobs added during October 2007 or October 2013. We can lose Full-time jobs. Will we record the addition of 500,000 jobs, as we did during 2016, or nearly 800,000 jobs as we did during 2015 or one million jobs as we did during 2014?
Could the Jobs Report report a net loss of jobs? The devil is in the seasonal factors. The seasonal factor for full-time jobs bounce around from year to year. The same can be said of the seasonal factor for the part-time jobs number. Thorough in those combinations and permutations and there is "no telling" where this number will land. Last month should have be an oddity created by three hurricanes hitting the United States.
Unemployment Should fall below the October 2006 level. The weekly unemployment claims data for October 14 was 1.597M claims. The weekly unemployment claims data for September 9, the date closest to the 12th of the month was 1.737 Million. Not all employees are covered by unemployment insurance. The percentage covered has been running from 1.2% to 1.4% of roughly 4.5%. A drop of 140,000 unemployed workers may equate to a drop of 400,000 U-3 unemployed workers. It would only take a drop of 280,000 workers to drop below the 2006 level.
How will a surge in people working multiple jobs impact the numbers? A cursory examination of the October data, both month to month and October to October, show that we could see over 8 million people working multiple jobs this October. We may exceed the level recorded last October (8.059M.) We may exceed the level recorded during October 1997 (8.139M.) Last month the drop in the MJH numbers allowed the workers number to fall while the jobs number actually improved. This month we may see some people participating more than others. The data indicates that we should see a comparable participation rate to last month, slightly higher or slightly lower than September. The participation rate should then improve through the end of the year before the New Year's Layoffs.
If we see a smaller drop in unemployed workers than we see in the gain of workers the workforce participation rate will improve. The question is "will the gains in real jobs be converted to seasonally adjusted losses? Will losses in unemployed workers remain a "negative." We could see participation record improve and have it reported as losses. Good news could become "bad" news. We could also see a drop in participation as more people work multiple jobs. Another item to watch when the "Jobs Report" is released is to see if some of the leisure and hospitality workers who lost jobs regained them and if we will continue to see a surge in construction jobs as the recovery from three hurricanes picks up speed. Will we see a surge in "senior citizen workers" and a drop in the U-7 unemployment rate?
It's the economy.