Written by Steven Hansen
We have had the weekend to ponder last Friday’s BLS Job report. I want to emphasize that there is little basis to understand jobs growth (except to use ADP’s numbers) because of the size of backward revisions to the BLS numbers. Consider the current vs original non-farm private jobs growth:
ADP’s backward revision is normally small. Today the Conference Board reported their July 2012 employment trends index rose marginally. This index is now saying employment growth will be slow in the coming months according to the index authors.
The Conference Board Employment Trends Index™(ETI) edged up in July, after declining the previous month. The index now stands at 108.11, up from the revised figure of 107.68 in June. The July figure is 5.9 percent higher than a year ago.
“The Employment Trends Index increased slightly in July, but is still below the May level, and is only slightly above the February level, suggesting that slow employment growth is likely to continue in the next few months,” said Gad Levanon, Director of Macroeconomic Research at The Conference Board. “There is no reason to expect employers to rapidly expand their workforce in the current economic environment, and the July pace of job growth (163,000) is unlikely to be sustained.”
July’s increase in the ETI was driven by positive contributions from five of the eight components. The improving indicators – beginning with the largest positive contributor – were Initial Claims for Unemployment Insurance, Industrial Production, Number of Temporary Employees, Real Manufacturing and Trade Sales and Percentage of Respondents Who Say They Find “Jobs Hard to Get.”
To add context to this index, the following graph compares BLS non-farm payrolls and the Econintersect Employment Index to the ETI. Econintersect uses non-labor and mostly non-monetary economic pulse points in constructing its index, while The Conference Board uses mostly elements of employment data.
Comparing BLS Non-Farm Employment YoY Improvement (blue line, left axis) with Econintersect Employment Index (red line, left axis) and The Conference Board ETI (yellow line, right axis)
The graph above offsets the Conference Board ETI by 3 months. My take is that neither Econintersect or The Conference Board ETI is mimicking the actual BLS jobs data. The Conference Board tries to predict turning points (which it appears to be good at) – but is unable to predict the intensity of the upward or downward movements. Econintersect attempts to predict intensity of movement (which it appears it is much better at than the ETI) – and is predicting a continuing weak labor market (which the labor market until recently has been stronger than Econintersect‘s index would have predicted).
One more kick at the employment situation using ADP vs BLS data.
Total Private Non-Farm Employment – ADP (blue line) vs BLS (red line)
You will note that BLS’s employment has been stronger then the ADP’s number since the beginning of 2011. I tend to trust the ADP number more in real time due the size of backward revision to the BLS numbers.
When the BLS tells you employment gains are xxxx, on what basis does your brain process this number? Is it from the number you heard last month, or the revised numbers this month?
Historical Unadjusted Private Non-Farm Jobs Growth Between June and July (Table B-1, data in thousands)
/images/bls non-adjusted change.PNG
The BLS numbers simply lack perspective, and are too heavily revised to be any use in real time. It does make good fun watching the market reacting to these numbers.
Caveats on the Employment Trends Index
According to the Conference Board:
The Employment Trends Index aggregates eight labor-market indicators, each of which has proven accurate in its own area. Aggregating individual indicators into a composite index filters out “noise” to show underlying trends more clearly.
The eight labor-market indicators aggregated into the Employment Trends Index include:
- Percentage of Respondents Who Say They Find “Jobs Hard to Get” (The Conference Board Consumer Confidence Survey
- Initial Claims for Unemployment Insurance (U.S. Department of Labor)
- Percentage of Firms With Positions Not Able to Fill Right Now (© National Federation of Independent Business Research Foundation)
- Number of Employees Hired by the Temporary-Help Industry (U.S. Bureau of Labor Statistics)
- Part-Time Workers for Economic Reasons (BLS)
- Job Openings (BLS)
- Industrial Production (Federal Reserve Board)
- Real Manufacturing and Trade Sales (U.S. Bureau of Economic Analysis)
Unfortunately many of these indices are not accurate in real time being subject to at times significant backward revision.