LIARS IN GOVERNMENT:
The divergence in the two data series, historically convergent, can be seen highlighted on the chart below:
While from a distance the highlighted area may not amount to much, here it is zoomed in just for 2013. The difference becomes quite pronounced, and amounts to just shy of 60K jobs per month on average for 2013 alone.
Putting the above into words:
Finally, the chart below shows that while until 2013 the divergence between two data series has been mostly cluster-free except for the Lehman collapse and the period just after it promptly normalizing thereafter, the past 7 months have seen a dramatic imbalance in data benefitting the algo-headline scanner moving NFP data, which on a 3 month trailing basis is almost as wide as it has been at any point in the past 5 years and just shy of the wides seens just after the Lehman collapse.
This means that either the JOLTS survey is substantially underrepresenting the net turnover of workers, or that once the part-time frenzy in the NFP data normalizes, the monthly job gains will plunge to just over 100K per month to "normalize" for what has been a very peculiar upward "drift" in the NFP "data."
And just like last month we will conclude with the same advice to the BLS: when manipulating data series across dimensions, make sure the manipulations foot across, and not just in 1 dimension.
Bureau of Labor Statistics Misrepresenting 2013 Job Gains By Over 40%
Many were surprised when last month we exposed the divergent lies at the Bureau of Labor Statistics when comparing two otherwise convergent data sets: the monthly all-important Non-Farm Payroll report and the (one month-delayed) JOLTS survey. Specifically, what we showed is that the Net Turnover from JOLTS (Hires less Separations) is now 40% below the trendline of cumulative job additions implied by the Non-Farm Payroll report's Establishment survey which has become the holy grail for both the stock market and the Federal Reserve's tapering ambitions. Following the release of the June JOLTS update, we can report that the divergence within BLS data series continues, and that the average monthly US job gain for the first 6 months of 2013 is either 198K if one uses the non-farm payroll data, or 30% lower, 140K to be specific, if one uses the JOLTS net turnover number.The divergence in the two data series, historically convergent, can be seen highlighted on the chart below:
While from a distance the highlighted area may not amount to much, here it is zoomed in just for 2013. The difference becomes quite pronounced, and amounts to just shy of 60K jobs per month on average for 2013 alone.
Putting the above into words:
- In April, according to JOLTS, there were 108K job additions. According to the NFP data, the job gain was 199K or 84% more than per JOLTS
- In May, according to JOLTS, there were 109K jobs additions. According to the NFP data, the job gain was 176K or 62% more than per JOLTS
- In June, according to JOLTS, there were 120K jobs additions. According to the NFP data, the job gain was 188K or 57% more than per JOLTS
- Adding across for all of 2013 (through the end of June data), JOLTS would have us know that only 837K jobs were added (or 140K per month average). Compare this to the 1,185K new jobs according to the Establishment Survey (198K per month average).
Finally, the chart below shows that while until 2013 the divergence between two data series has been mostly cluster-free except for the Lehman collapse and the period just after it promptly normalizing thereafter, the past 7 months have seen a dramatic imbalance in data benefitting the algo-headline scanner moving NFP data, which on a 3 month trailing basis is almost as wide as it has been at any point in the past 5 years and just shy of the wides seens just after the Lehman collapse.
This means that either the JOLTS survey is substantially underrepresenting the net turnover of workers, or that once the part-time frenzy in the NFP data normalizes, the monthly job gains will plunge to just over 100K per month to "normalize" for what has been a very peculiar upward "drift" in the NFP "data."
And just like last month we will conclude with the same advice to the BLS: when manipulating data series across dimensions, make sure the manipulations foot across, and not just in 1 dimension.
No comments:
Post a Comment