Even a week after the January 2012 jobs report was released, pundits are still arguing over what it means. I too have been mulling over the data. The problem is that:
- There is significant controversy about the way the Bureau of Labor Statistics (BLS) seasonally adjusts the data. This analyst believes the adjustment methodology is way too complex, and initially it is applied to incomplete data which is subject to significant revision for months after the first release, followed by annual revisions. Whatever the reason, the end result is a jagged looking seasonally adjusted data – and from a systems point of view, seasonally adjusting data is supposed to provide a smoothing of the curve to provide clarity out of chaos. By definition of the intent of seasonal adjustment, the BLS seasonal adjustment methodology is broken. A good example of very volatile unadjusted data turned into a smooth interpretable seasonally adjusted data is the DOL’s Weekly Initial Unemployment Claims 4 Week Moving Average.
- Januarys are strange months. In reality, the non-farm workforce always contracts. over the last decade – the jobs market contracted at least 2.5 million each year between December and January. Seasonal adjustments add at least 2.5 million people to the January jobs count. If the methodology even has a little error, the January jobs numbers are garbage. IMO, this is why the muted response on the good jobs numbers by those who drill into employment statistics.
- To add to the confusion, any person with their head screwed on straight can point out the inconsistencies in the BLS jobs report. The Household and Establishment portions of the report are done using different methodologies, use different measuring sticks, have different sampling uncertainties – and, as you could guess, have different answers.
It is the job of analysts to make sense out of inconsistency. Real world data is messy.
More context to the January jobs report can be provided using the BLS’s data. The graphic below shows jobs growth defined by using the BLS seasonally adjusted data and their unadjusted data.
The above graphic should be illuminating because of the smoothness of the curve (blue bars). This is the way seasonally adjusted data is supposed to look – yet we are seeing unadjusted data. Both the adjusted and unadjusted databases have the same area under the curve (same jobs growth over time).
Using the unadjusted BLS non-farm data’s year-over-year job growth divided by 12, we can see that monthly jobs growth over the last 12 months is in the zone of 160,000 – exceeding the jobs growth necessary for population growth for the last 12 months.
Jobs growth is at the highest level since the end of the 2007-09 recession. Using this method the average monthly jobs growth at 161,000 for the last twelve months, or the year-over-year monthly jobs growth, improved 9.6% over last month. I can be very sure the error in this statement is small. Other takeaways:
- Historically, recessions do not start with employment rate of growth expanding
- Rate of employment growth expanding historically indicates a strengthening economy
- The total non-farm job level is roughly at January 2005 levels – yet the USA population has grown by 18,100,000.
I find it dishonest to say that non-farm jobs grew 243,000 as headlined by the BLS when jobs in reality contracted over 2,500,000, just like they do every year. As an old boss once told me “say what you mean, and mean what you say”.
Economic News this Week:
The Econintersect economic forecast for February 2012 was released this week, and continues to indicate a strengthening of economic fundamentals. This index essentially uses non-monetary measures (counting things) to determine economic growth or contraction. Many of this index’s components draw on transport industry movements.
ECRI has called a recession. Their data looks ahead at least 6 months and the bottom line for them is that a recession is a certainty. The size and depth is unknown but the recession was to hit before the end of 1Q2012. At this point, the chance of a recession based solely on the coincident data is becoming unlikely.
This week ECRI’s WLI index value was -4.3 – a negative value but the best index value since August 2011. This is the fourth week of index value improvement. This index is indicating the economy six months from today will be weaker – but increasingly marginally.
Initial unemployment claims fell 15,000 to 358,000. Historically, claims exceeding 400,000 per week usually occur when employment gains are less than the workforce growth, resulting in an increasing unemployment rate (background here and here). The real gauge – the 4 week moving average – fell dramatically to 366,250. Because of the noise (week-to-week movements from abnormal events AND the backward revisions to previous weeks releases), the 4-week average remains the reliable gauge.
The data released this week which contains economically intuitive components (forward looking) were positive – and are trade imports, JOLTS job openings, and rail car loadings.
Weekly Economic Release Scorecard:
Bankruptcies this Week: Global Aviation Holdings, TBS International
Failed Banks this Week: