JOHN WILLIAMS’ SHADOW GOVERNMENT STATISTICS

FLASH UPDATE

March 6, 2009

 __________

February Payroll Loss Was 899,000
Net of Concurrent Seasonal Factor Bias

Annual Percentage Contraction in Jobs at 50-Year Low

SGS-Alternate Unemployment Rate at 19.1%

__________

PLEASE NOTE: An Alert is planned for over this weekend to cover: ongoing developments in the still-unfolding systemic solvency crisis; what appears to be a sudden slowing in broad money growth (February); a review of why the U.S. dollar has been relatively strong despite miserable and deteriorating underlying fundamentals; and where the U.S. currency is headed over the longer-term (the outlook for a massive decline is unchanged).

The next regular Flash Update is planned following the release of the February retail sales report on Thursday (March 12th). The next full newsletter should follow early in the week of March 15th. Any intervening Flash Updates or Alerts would be published as dictated by economic or financial developments. 

– Best wishes to all, John Williams

Concurrent Seasonal Factors Go Awry. This morning’s (March 6th) reported February jobs loss of 651,000 was suspiciously close to revamped consensus expectations of a 650,000 loss, yet the loss would have been 812,000 without revisions to January’s level, and would have been 899,000 net of the Concurrent Seasonal Factor Bias (CSFB), as also discussed in a separate section below.

Where month-after-month, the prior month’s reporting consistently has been revised much lower, along with the reporting of latest monthly headline payroll number, what is at work here, is more than just an initial overestimation of the seasonally-adjusted headline numbers, it involves monthly revisions to the concurrent seasonal factor adjustments. The concurrent adjustments allow for unusual monthly revisions that would not have been possible in earlier years, when monthly seasonal factors were set in advance for at least six months at a time.

Consider the latest payroll reporting. The 651,000 seasonally-adjusted jobs loss reflected a drop in reported payroll employment level to 133.768 million in February, from a revised 134.419 million in January. January, however, initially had been reported at 134.580 million, and had that not been revised, February’s seasonally-adjusted monthly decline would have been 812,000, instead of the market-soothing, near-consensus 651,000.

So, where did the 161,000 downward revision to the January employment level come from? One would expect the change was in raw data from more complete information, but the rawest data published for January — the not-seasonally-adjusted series — revised upward minimally by just 6,000, from 132.341 million to 132.347 million, suggesting that the January revisions were tied primarily to monthly revisions in the seasonal-adjustment factors.

Nonfarm Unemployment May Have Hit 34% During Great Depression. While the official estimate of peak total unemployment during the Great Depression was 25% in 1933, such was in an economy where agricultural employment accounted for roughly 27.7% of the total (1930 federal census). Nonfarm unemployment has been estimated for 1933 at about 34% (Social Security Administration). Where today’s nonfarm employment (household survey) accounts for roughly 1.4% of the total, with nonfarm employment at 98.6%, the 1933 nonfarm unemployment estimate might be the more appropriate number for comparison purposes with today’s economy. 

Keep in mind, though, that the Great Depression unemployment rates were estimated after the fact, without adequate data or surveying. Regular unemployment surveying did not start until the early 1940s and then was subjected to methodological revisions over the decades. As discussed below, the SGS-Alternate unemployment rate now tops 19%, and such would be my best estimate of a rate that would be comparable to the Great Depression readings.  

Latest Reporting Showed Rapidly Sinking Economic Activity. Putting aside any games the Bureau of Labor Statistics (BLS) may be playing with the data, the official numbers indicate broad economic activity still is sinking rapidly.

Payroll Survey. The BLS reported a statistically-significant, seasonally-adjusted jobs loss of 651,000 (down 812,000 net of revisions) +/- 129,000 (95% confidence interval) for February 2009, following a revised 655,000 (previously 598,000) jobs loss in January. Annual contraction (unadjusted) in total nonfarm payrolls continued to deepen, down 3.12% in February, versus a revised 2.57% (previously 2.58%) in January. The annual decline in February was the deepest in 50 years, since August 1958. The seasonally-adjusted series also continued contracting year-to-year, down by 3.02% in February, versus a revised 2.65% (was 2.53%) contraction in January.

Concurrent Seasonal Factor Bias. The pattern of impossible biases being built into the headline monthly payroll employment continued with February 2009 reporting (see the opening section above and the Reporting/Market Focus in SGS Newsletter No. 43 of June 10, 2008). Instead of the headline jobs loss of 651,000, consistent application of seasonal-adjustment factors — net of what I call the concurrent seasonal factor bias — would have shown a more-severe monthly jobs loss of about 899,000. This upside reporting bias has been seen in 11 of the last 12 months, with a rolling 12-month total upside headline-number bias of 1,325,000. A worksheet on this is available upon request.

Birth-Death/Bias Factor Adjustment. An element that may have helped to soften the reported February jobs loss was the monthly bias factor (birth-death model). Never designed to handle the downside pressures from a recession, the model adds a fairly consistent upside bias to the payroll levels each year, but with a seasonal distribution that has one month of negative bias: January. The downside adjustment to unadjusted January 2009 payrolls was 356,000, but the February bias moved into positive territory, a monthly add of 134,000, a net month-to-month positive swing of 490,000 jobs.

Household Survey.  The usually statistically-sounder household survey, which counts the number of people with jobs, as opposed to the payroll survey that counts the number of jobs (including multiple job holders), showed February employment down by 351,000, where January had been down by roughly 832,000 adjusted for population changes.

The February 2009 seasonally-adjusted U.3 unemployment rate showed still another statistically-significant increase, to 8.08% +/- 0.23%, from 7.56% in January.  Unadjusted U.3 rose to 8.9% in February from 8.5% in January.  The broader February U.6 unemployment rate jumped to an adjusted 14.8% (16.0% unadjusted) from 13.9% (15.4% unadjusted) in January.

During the Clinton Administration, "discouraged workers" — those who had given up looking for a job because there were no jobs to be had — were redefined so as to be counted only if they had been "discouraged" for less than a year. This time qualification defined away the bulk of the discouraged workers. Adding them back into the total unemployed, actual unemployment, as estimated by the SGS-Alternate Unemployment Measure, rose to about 19.1% in February, from 18.0% in January.

Employment Environment. The continued broad deterioration in February’s employment environment broadly was in line with deterioration in the better-quality employment-environment indicators: January newspaper help-wanted advertising hit another successive historic low, with a deepening annual fall-off in February online help-wanted advertising (Conference Board); new claims for unemployment insurance have continued to surge sharply; and employment readings continued in the deepest recession territories for both the February manufacturing and nonmanufacturing purchasing managers surveys.

Week Ahead: Retail Sales: Expectations (Briefing.com) are for roughly a 0.4% monthly decline in seasonally-adjusted February retail sales, due for release next Thursday (March 12th). Net of inflation-adjustment, the series likely contracted on a monthly basis, with the pace of annual contraction deepening to a new low for the historical series dating back to the period following World War II.

Trade Deficit: While the markets may be looking for a slight narrowing of January’s monthly trade deficit, due for release next Friday (March 13th), recent heavy reporting distortions and likely revisions allow for a significant deterioration.

__________