The Flawed BLS Jobs Report

by: Steven Hansen

Over the last months, I have resolved not to deeply analyze the monthly release of the Bureau of Labor Statistics (BLS) employment data. I believe the data is flawed to the point MoM data analysis has an unacceptably high risk of coming to the wrong conclusion.

Over the last year in my analysis, I was sidetracked into pointing out the distortions in the data – and losing sight of the big picture..

My objective is to get a good handle on economic activity. We do not need BLS data to analyze employment MoM. We have ADP data which is arguably better groomed for MoM analysis than the BLS data but is unfortunately presented in significantly less detail.

The BLS Household Survey which produces the unemployment rate jumps up and down like a yo-yo. The fact is that when your system shows the unemployment rate falling while fewer and fewer people have jobs – you have a report giving you the wrong answer. It is illogical, and it has happened several times to the unemployment rate during our Great Recession.

Click to enlarge

The real magic in the unemployment numbers comes from the ability of the BLS to shrink the workforce. It is easy to pretend someone who does not work, cannot find a job, and has exhausted getting benefits is no longer a bona fide card carrying member of the American workforce.

Click to enlarge

As evidenced by the rate of change in the workforce curve beginning in 2008, is it not strange that so many people decided not to work in such a short period?

All the necessary data is published in enough detail by the BLS to perform an analysis – it is just buried well below the headline data. Are the villains the media who ignore the real data? It is easy to blame the media as they generally are as lazy as they are ignorant.

Understand that the BLS spoon feeds the headlines to the media. If they wanted the public to really understand what is going on they would headline with the data they bury.

The establishment survey data of the BLS is too erratic for accurate month-to-month evaluation. Two graphic examples using data from ADP as a control on the BLS data follow:

Click to enlarge

This first graph illustrates an obvious methodology change in the BLS data. No explanation is ever given in the BLS analysis on why their number had a change in rate..

Click to enlarge

There is an obvious difference in the ADP smoothed data to the BLS jerky data. All the data I am analyzing here is adjusted. The above graph compares the employment data for service industry. When you look at just released data and compare it MoM, what accurate conclusions can you draw from jerky data (the big boys call this “noise”)?

This is why I do not like to review the monthly BLS jobs report. It does not matter how good their methodology looks on a piece of paper - it is results that count and their data is too erratic for proper evaluation.

This does not mean I ignore BLS data. The BLS data is useful (not to mention has considerable detail) when you use it over longer periods of time where the monthly distortions average out. An example is the graph below which compares US Census population numbers to BLS non-farm employment.

Click to enlarge

Using proportioned scales, this chart shows visually that employment to population ratios are well under 1990 levels. It also makes my point that for unemployment rate to truly improve, employment must grow faster than the rate of population increase.

And to get a good understanding of how terrible the jobs creation picture looks, we can get this data from BLS's JOLTS program.

Click to enlarge

This data confirms we are not creating jobs.

Is the BLS data manipulated? When you adjust definitions or use totally asinine methodology – that is manipulating data. Politicians of both parties are guilty of orchestrating the change in definition of who is unemployed throughout the years.

When you present data in a way that it will be misinterpreted – that is manipulation . Ask whether the data presentation would be different if high unemployment numbers were considered a good thing. All data is spun for the good of the entity which issues it.

But, in the case of the BLS, their product is as flawed as brakes on a Prius. They have several methodology issues which needs attention yesterday.

The service the BLS should be performing for the country is continuing to put forward studies on employment. In spite of all the criticism thrown at the Federal Reserve, they continuously critique their purview through the publishing of studies and reports.

Why is employment analysis left to the non-government sector? Clusterstock published a provocative graphic on the amount of researchers by region, and their relative change over time. Innovation is a function of research. Jobs are a function of innovation.

Click to enlarge

The government through the Department of Labor and the BLS needs to be front-running efforts to improve the employment picture.

The weekly initial unemployment claims improved slightly WoW.

Click to enlarge

This claims data is a reversal of last weeks unexpected jump, but we should not be seeing these high job loss numbers over two years after the start of our Great Recession. As several readers have pointed out, part of the improvement in the claims numbers might be attributed to the bad weather preventing new claimants from filing (hat tip to ain't no fortunate son and John Lounsbury).

This recession continues to be as much about jobs as it is about debt.

Manufacturing and Business

This week I would like to introduce a new report now being released - the Ceridian-UCLA Pulse of Commerce Index™ .

The index is based on an analysis of real-time diesel fuel consumption data from over the road trucking tracked by Ceridian Corporation. Ceridian produces the index in conjunction with UCLA Anderson School of Management and Charles River Associates. It mirrors closely the Federal Reserve's Industrial Production Index but is issued days before that index is released.

The index is built by analyzing Ceridian's electronic card payment data that captures the location and volume of diesel fuel being purchased by over the road trucking operations, providing a detailed picture of the movement of products across the United States. "Goods have to be transported for an economy to grow, so it will be important to monitor this index to see if the economy is really on the move.

I have been trying for a way to measure diesel fuel usage as a check on economic activity. With the addition of this new report into the stable of indicators I follow, I believe we can now completely validate the economic data that is being released. The highlights of the report:

….. the U.S. economy fell in January after a significant increase in December, with the index falling at an annualized rate of 36.8 percent. The more reliable three-month moving average for January managed to show a 3.3 percent gain at an annualized rate following the exceptional annualized rate of 14.6 percent in the previous month.

"Though the January 2010 number is disappointing, the index is 3.6 percent above its January 2009 level and is similar to year-over-year pre-recession values," said Edward Leamer, director of the UCLA Anderson Forecast and chief economist for the Ceridian-UCLA Pulse of Commerce Index. "Also, the three-month moving average is 2.3 percent above the previous year's value, which is the first time that there has been a year-over-year increase since April 2008, 21 very difficult months ago."

The latest PCI numbers suggest caution about celebrating the recently announced 5.7 percent GDP growth number. Although the 7.3 percent growth rate in the fourth quarter of 2009 for the PCI was strong, at that rate the index won't exceed the 2007 second quarter peak until the third quarter of 2011. "Things are going to have to look a lot better in February and March to turn this worry into optimism about the power of the recovery," Leamer said. "Stay tuned. We expect this showing in January indicated by the PCI will also be seen in the Industrial Production number when it is released later this month."

Click to enlarge

As our economy stabilizes, imports continue to grow faster than exports as the trade deficit increased nearly $4 billion (pdf). The growth in exports was nearly uniform across all categories, while the import growth was concentrated in oil products and foods.

Click to enlarge

The government released December 2009 manufacturing and trade sales and inventory. Their statement:

The U.S. Census Bureau announced today that the combined value of distributive trade sales and manufacturers’ shipments for December, adjusted for seasonal and trading-day differences but not for price changes, was estimated at $1,040.4 billion, up 0.9% (±0.3%) from November 2009 and up 4.7% (±0.5%) from December 2008. Manufacturers’ and trade inventories, adjusted for seasonal variations but not for price changes, were estimated at an end-of-month level of $1,310.7 billion, down 0.2% (±0.1%) from November 2009 and down 9.7% (±0.4%) from December 2008. The total business inventories/sales ratio based on seasonally adjusted data at the end of December was 1.26. The December 2008 ratio was 1.46.

My take is that seasonal adjustments are too varied for any degree of certainty that manufacturing sales are improving. Of all months, December appears the most uniform – and it is clear sales are up YoY but less than 2006 levels. We need at least one more month of YoY increases to begin to think there is an expansion cycle beginning. Looking from a glass half full perspective, there is no evidence of a decline in the November and December data.

From the same data set, I extracted business sales which clearly demonstrates expansion.

Click to enlarge

Advance data from another subset of this series, retail and food services was issued for January 2010 and headlined as follows:

Advance estimates of U.S. retail and food services sales for January, adjusted for seasonal variation and holiday and trading-day differences, but not for price changes, were $355.8 billion, an increase of 0.5% (±0.5%)* from the previous month and 4.7% (±0.5%) above January 2009. Total sales for the November 2009 through January 2010 period were up 4.3% (±0.3%) from the same period a year ago. The November to December 2009 percent change was revised from -0.3% (±0.5%)* to -0.1% (±0.2%)*. Retail trade sales were up 0.5% (±0.5%)* from December 2009 and 5.3% (±0.5%) above last year. Gasoline stations sales were up 29.0 % (±1.5%) from January 2009 and non-store retailers sales were up 12.4% (±1.7%) from last year.

My take is that the advance data is not demonstrating enough of improvement YoY, and calls into question what retail expansion is occurring.

Click to enlarge

Other Data

Federal Reserve Chairman Ben Bernanke revealed the mechanics for soaking up liquidity in the system if the economy starts moving. One of the mechanics involved selling large blocks of term deposits at a higher rate than the Federal Funds rate – thus allowing the Federal funds rate (now from zero to 1/4%) to be adjusted independently.

His speech was smoke and mirrors concerning how he was going to unwind the massive Fed balance sheet. His statement pretty much left all doors open.

However, to help reduce the size of our balance sheet and the quantity of reserves, we are allowing agency debt and MBS to run off as they mature or are prepaid. The Federal Reserve is currently rolling over all maturing Treasury securities, but in the future it may choose not to do so in all cases. In the long run, the Federal Reserve anticipates that its balance sheet will shrink toward more historically normal levels and that most or all of its security holdings will be Treasury securities. Although passively redeeming agency debt and MBS as they mature or are prepaid will move us in that direction, the Federal Reserve may also choose to sell securities in the future when the economic recovery is sufficiently advanced and the FOMC has determined that the associated financial tightening is warranted. Any such sales would be at a gradual pace, would be clearly communicated to market participants, and would entail appropriate consideration of economic conditions.

The markets did not react to the statement. My surprise was that Bernanke liquidity drying methodology is so clearly aimed at taking lending money out of the system. How does this square with the politicians pressure on banking to lend?

The weekly Mortgage Bankers Association new mortgage application data for the week ending 05 February 2010 decreased slightly but remains about 60% of the level of early 2009 using seasonally adjusted data. The 30 year fixed mortgage rate deceased 7 basis point to 4.94%.

Bankruptcies this week: Penton Media

Failed Banks This Week:


Economic Forecasts Published this Past Week

The Economic Cycle Research Institute (ECRI) released their Weekly Leading Index declining for nine straight weeks. Lakshman Achuthan, Managing Director at ECRI added:

While the U.S. economic expansion is well set to strengthen in the near term, the current easing in WLI growth to a 25-week low suggests that growth will begin to throttle back by mid-year.

This week the President published his 462 page (284 pages before references, tables and appendixes) 2010 Economic Report (pdf). I scanned the document looking for bits of interest. The document lays out the government's view of the crisis. and details the government's actions taken in their attempt to mitigate the effects of the Great Recession.

Overall I thought the presentation was good overview of the financial benefits of the programs this administration would like to embark – education, health care, global warming, green jobs.

I admit I did not read every word. It did leave me wondering who actually reads this as the information is mostly a repeat for the economically informed – and too boring for those who are not economically informed..

Click to enlarge

The Philly Fed released the results of their 1Q 2010 Survey of Professional forecasters.

Click to enlarge

Disclosure: GLD, XPH, XLB, XLV, FTR, IOO, HYG, KSWS, PIN, XHB, WMT, GDX, Physical Gold - as well as numerous puts and calls which comprise less than 3% of my portfolio.