Reclaiming Common Sense

December Jobs Report Could Be Big Present, Little Present, or a Seasoned Turkey


The Monthly Jobs Report, or Employment Situation Report, contains a cornucopia of information. It is created using the Current Employment Statistics (CES) data on workers and the Current Population Survey (CPS) data on jobs. The data that is recorded is the non-seasonally adjusted data. The data that is reported is the seasonally adjusted data. What should we expect to be recorded and what should we expect to be reported for the December Jobs Report, if it is released?


Why wouldn't the jobs report be released? The November New Home Sales data was supposed to be released the Thursday prior to Christmas. It was not released, due to the government shutdown. It is possible, not probable, that the jobs report will not be released for the same reason.


Fortunately, the ADP Payroll report will be released the Thursday before the scheduled Jobs Report. The Article "December ADP Report Should Be Jolly" used similar techniques to analyze the seasonally adjusted data. There is not a "non-seasonally adjusted data set" for the ADP data that is publicly available. It was projected that based on the current year data that we should add 201,000 to 220,000 payroll positions. The Month to Month data indicates that "Information" will trim payroll. We could see "Other Services" and "Natural Resources" trim their payrolls. The Largest gains should be in Professional Business Services (PBS,) Education and Health Services (EHS,) Leisure and Hospitality (LAH,) Other Services (OS,)  and Natural Resources (NR.) The month to month growth projects 204,000 to 268,000. Expect a number around 230,000.


The December to December ADP data indicated that we should expect the largest gains in  PBS, NR, "Construction," EHS, and LAH. We should also expect a number between 218,000 and 256,000 or roughly  number around 237,000.


Throwing Darts for calculating the data. The larger the seasonal factor the higher the possible SA CES data to be reported. One of the highest seasonal factors used was used during December 2016. One of the lower seasonal factors (SF) used since 2000 was used during 2017. If the 2016 SF was applied to the 2017 data then we would have had 240,000 workers reported as being added instead of 148,000. If the 2017 SF was used for the 2016 report then the 252,000 workers added would have been reported as 161,000.


The Current Employment Statistics data is closest to the ADP Payroll data. The "Headline" jobs number is the Non-Farm payroll data, which includes the Government Worker Data. The ADP Payroll data does not include government worker/payroll data and is closest to the "Total Private Sector Job. Unlike the seasonally adjusted (SA) ADP data, there is a remarkable range for potential December to December and November to December growth. The pace of of growth has been approaching 2.00% this year, even after the substantial revisions made to the non-seasonally adjusted (NSA) and SA data between the December 2017 and January 2018 Jobs report. There is the possibility for a strong December kick to the data. If we grow at a rate of 1.95% then we could lose NSA CES workers and still add 208,000 private sector workers. We should also add SA government workers, so this number should be reported even higher than 208,000. If we grow at 1.98%, tied for the best that we have this year, that number hits 252,000. This method projects a number around  235,000.


The Current Year data is trending with 2005, 2013 and 2015. Contrary to what you have read elsewhere, this has been a solid year for NSA CES worker additions. If we trend with 2015 then we could add 94,000 NSA CES workers. If we trend with 2005 then we could trim 92,000 NSA CES workers. Split the difference and we have "no change." We could also trim 116,000 as we did during 2013, which is comparable to 2003 and 2004. We could go from only adding 110,000 workers, using the 2017 SF, to as high as 425,000 using the 2016 SF.  This method is projecting somewhere around  234,000 SA private Sector Workers.


The Month to month data paints a similar picture. We could see month to month contraction or expansion. It seems obvious to the casual observer. Some months we almost certainly add workers while others, especially January, we shed workers. We could drop by 0.10% or increase our workers by 0.10%.  It could even go "unchanged." If we add no workers, NSA, the SA could be reported between 201,000 and 326,000 private sector workers. If we shed workers as we did during 2016 and 2017we could add between 115,000 and 252,000 workers, depending upon the SF used.  The bulls eye is at 234,000 again.


The Devil is in the revisions. The non-farm payroll revisions are published on the employment situation website. The "final" September revision was upward 175,000 workers higher than the "advance value." That was a good month under former President Obama. There have been fairly regular revisions to the NSA CES NFP data of -8,000 to 30,000 workers with the first revision to the data, in this case the revisions to the November data. If the November data is revised down 8,000 that will effectively boost this month's data by 8,000. A 234,000 number would be reported as 242,000. If November is revised higher by 30,000 then this month's 234,000 will be reported as just 204,000. Also remember that the CES data will be revised between the release of this report and the January report.


Which sectors will grow and which will be trimmed month to month and December to December?  The month to month data indicates that we will record  gains, NSA, only in Manufacturing, Trade, Transportation, and Utilities (TTU,) and Financial Activities (Fin. Act.) The only sector where we should see a drop reported, SA, is in IT. The largest SA CES gains month to month should be Construction, MFG, PBS, EHS, and LAH. We could see a drop in IT and in TTU reported. The December to December data indicates that all sectors should record increases except IT, with the largest gains in Construction, PBS, EHS, and LAH.


What about Jobs? We almost always lose jobs during December. There are a number of Scrooges in this world in which we live. Expect to see NSA full-time (FT) jobs and NSA Part-time (PT) jobs record a drop. It is possible that we could see NSA PT jobs climb as they did prior to the "Great Recession." The seasonally adjusted data will almost certainly convert these losses into gains for both FT and PT jobs. The problem with forecasting the jobs data is that the full-time job seasonal factor is different from the part-time job seasonal factor. These factors change month to month and year to year.


What about Unemployment? Both the NSA and SA U-3 unemployment numbers should fall this month. If the SA Jobs number is enough of a positive number, then the negative unemployment number will mean that participation should improve from last December's level. The December Participation rate, NSA, had languished at roughly 62.40%. the unemployment rate, is the lowest that it has been since at least 1981. Unemployed workers are participants. The unemployment rate was falling under former President Obama because the participation rate was falling, not the other way around. Once again, it is difficult to project the SA U-3 unemployment value because the seasonal factor used for the U-3 unemployment value is different from the FT seasonal factor and from the PT seasonal factor. The seasonally adjusted participation rate is the seasonally adjust U3 plus the SA FT plus the SA PT number all divided by the workforce population. There are too many moving parts to accurate project a seasonally adjusted unemployment rate or a seasonally adjusted participation rate.


How is this President ding compared to his predecessors? This column has referenced the "Four Presidents at 96 months" article for nearly two years. The article detailed how the participation rate for January 2017 could have been reported lower if the workforce population had grown during that month. last month we had a record level of November full-time workers. "Five Presidents at 22 months" details the changes we have seen. We have more FT Jobs than we have ever had during this time of year. Full-time jobs tend to pay better than part-time jobs. What will this month's "Five Presidents" article reveal?


What is going to happen with regard to the multiple job workers? This column has proposed that employers were hiring based upon the hourly requirements of the Patient Protection and Affordable Care Act. Full-time equivalent hours mattered. We are seeing employer provided health care rise for the first time since the introduction of "Obamacare."  Multiple Job Workers Could rise this month. We will probably see a surge in people working two part-time jobs. There is the potential for a surge in people working a full-time job and a part-time job. This is part of the reason why unemployment is expected to fall. A part-time job should pay better than weekly unemployment benefits. Unemployment numbers are "unchanged" when a person working multiple jobs loses one job. A second job is the "ultimate" form of unemployment insurance. We may see another record for Women in the Workforce because women tend to work more dual part-time jobs than men work.


How are wages performing? We are seeing wage inflation.There are some prognosticators that are lamenting the difference between real wage growth and nominal wage growth. Where was this discussion when we had wage stagflation and deflation? We are seeing service inflation and commodity deflation. Could we be seeing service inflation because we have more service workers? Wages have been growing faster than the rate of inflation. This wage information will be examined in the "Wages and Workers" article.


But, what about the number of job openings versus the number of unemployed workers? This column has written numerous articles regarding how disingenuous it is to compare the JOLTS job openings data with the unemployment rate or level. The JOLTS data is comparable to the CES worker data. The Unemployment level is obtained from the CPS jobs data. They are different surveys with different sample sizes, different error rates, measuring different things. We might as well compare the job openings with the number of people collecting continuing unemployment benefits. Apples, Red Pears, Pomegranates. The data that is supposed to be released this week is the December Jobs Report and next week we are supposed to receive the November JOLTS report.


This jobs report projection is rather like throwing darts at a moving dartboard. You may be accurate, or you may be precise, or you may be lucky to hit the board. The overall indications are that we record a gain or loss of workers, and that we will see an addition f workers reported. Will it be 230,000? That depends upon the seasonal factors used and the adjustments to last month's numbers. Will we see wages increase again? Probably. We know that different sectors have seen differences in wage growth month to month. Even though the continuing unemployment claims data has risen since mid November, more people will have been looking for seasonal work, part-time work, or an extra job. 


Expect the unexpected. Did Santa bring you exactly what you wanted? Did Santa bring you socks and underwear because that is what you needed? Expect the unexpected. Smile for the camera even if it isn't the right size. Dig into the data and you might find the wishbone of the turkey and get your wish. It could be a large present, reported over 300,000, or a small present under 150,000, or a seasoned turkey around 230,000. It all depends on the growth rate, the seasonal factors, and the revisions.


It's the economy.