Tuesday, November 3, 2015

I hope everyone is expecting a big job gains number for October.

I hope everyone is expecting a big job gains number for October.  BLS has some making up to do.  If you look at the above graph you will see multiple years of raw, non-seasonally adjusted employment data plotted.  This data was downloaded from the BLS website.

Just looking at it you can see not much has changed in terms of the employment trend compared to 2014.   Yet this year as rate hikes seemed imminent because employment was strengthening suddenly the seasonally adjusted jobs gains reported by BLS nosedived.  The 2014 average reported seasonally adjusted gain was about 240k/month, while 288k/month was reported from Sept 2014 to Feb 2015.  But then suddenly March 2015 was reported at 126k, May, June and July rebounded but still all but June were below 230k.  Then August and September were 173k and 142k respectively. Since March seasonally adjusted gains have averaged just 197k/month while the comparable period in 2014 was 242k.

If you look at YoY calculations on the raw data, every month since June 2014 has been showing a gain of 1.94%-2.39% which is about 240k-260k/month.  If you average employment for the last 12 months and compare to the same period a year earlier you come up with job gains averaging 251k/month for 2015 plus the last 3 months of 2014.

Yes if looks like monthly job gains have slowed from a peak of around 260k/month to about 230k/month now.  However this slowdown has been drastically overstated by the BLS seasonally adjusted data which is showing gains well below 200k/month lately.  It appears the BLS seasonal adjusted data is 20%+ low for the year.

Another point of interest is that in 2014 BLS seasonally adjusted reports showed average monthly gains were about 240k/month while YoY was around 220k regardless if you looked at the months individually or the year as a whole.  The BLS seasonal adjustment was about 10% high on average and 20%+ high in many months.  While in 2015 suddenly the seasonal adjusted reading is on average about 20% low, and misses low by 30% in many months.

I'm not saying any of the new jobs created or good or high paying--the data clearly indicates otherwise--, just that it seems the quantity of jobs is understated by a wide margin this year.

1 comment: