Flash Update
FLASH UPDATE - July 11, 2007
JOHN WILLIAMS' SHADOW GOVERNMENT STATISTICS
FLASH UPDATE
July 11, 2007
__________
Revision and Seasonal-Adjustment Games Help Obfuscate Employment Reality
__________
The SGS Ongoing M3 Data Series has been updated for June's preliminary estimate, based on 25 days worth of data. The full-month estimate will be posted this coming weekend and discussed in the newsletter. The June/July newsletter is planned now for posting on July 16th, along with the Hyperinflation Summary. All postings will be advised by e-mail. -- Best wishes to all -- John Williams
Consistent Seasonals Suggest 107,000 June Jobs Gain (Not
132,000)
To maintain the reported 2,000,000 (exact) annual jobs growth in place for revised seasonally-adjusted May 2007 payrolls, monthly jobs gain reporting needs to be targeted by the Bush Administration at 167,000 per month. In contrast, the Clinton Administration targeted 250,000 per month (3,000,000 per year) for an extended period of time. Perhaps for fear of rattling the credit markets, however, initial monthly reporting typically has been "understated" over the last year or so, followed in subsequent months by major upward revisions to prior reporting. If this pattern is not a machination of the Bureau of Labor Statistics (BLS), and if the later numbers are accurate, not fabrications, the BLS would do well to suspend its initial reporting of these numbers, rather than to continue misleading the public and the markets with such poor quality reporting.
Massive prior-period revisions and unusual seasonal-adjustments distorted the June employment report. Seasonally-adjusted June payrolls were reported up by 132,000 (207,000 net of revisions) +/- 129,000 (95% confidence interval), following a revised 190,000 (previously 157,000) gain in May. With the BLS re-jiggering its seasonal factors each month, the range of potentially gimmicked reporting is great. Nonetheless, year-to-year change in the seasonally-adjusted and unadjusted series should be the same, and based on June's unadjusted annual growth rate, the jobs increase should have been 107,000 instead of 132,000. Annual growth for June was reported at 1.5% seasonally adjusted, 1.4% unadjusted.
Adding to the malleability of the reporting, June's fudge or bias factor (birth/death model) created 156,000 jobs that were added to the unadjusted number for good measure. That number flows through to the seasonally-adjusted data with minimal variation. The bias factors never are adjusted for a recessionary environment, always assuming ongoing growth. For example, 25,000 construction jobs were added to employment last June, and 26,000 construction jobs were fudged this June, despite the obvious deterioration in the 2007 construction environment.
Further, the reported 16,500 June jobs gain in the educational sector suggests that, as usually is the case, the BLS had problems properly seasonally adjusting for the effect of the end of the school year.
The household survey counts those who are employed, or not, while the payroll survey counts the number of jobs. Even allowing for the reported 7.5 million individuals holding multiple jobs, the two series are irreconcilable by the BLS. The household survey showed seasonally-adjusted June employment up by 197,000, with seasonally-adjusted U.3 unemployment increasing to 4.53% +/- 0.2% in June, from 4.46% in May. Unadjusted U.3 rose to 4.7% in June from 4.3% in May.
The broader, seasonally-adjusted U.6 unemployment rate held at 8.2% in June, but rose on an unadjusted basis to 8.5% from 7.9% in May. Net of the "discouraged workers" defined out of existence during the Clinton Administration, traditional unemployment continues to run around 12%.
Other economic reporting last week was light, except for some continuing strength in the purchasing managers surveys, which will be discussed in detail in the upcoming newsletter.
Ongoing M3 Growth at 13% for June. Based on 25 days of reporting in June and the major revisions to small time deposits (M2) and large time deposits (M3) discussed in the July 1st Flash Update, year-to-year growth for June M3 is estimated at 13.0%, down from 13.1% in May. June M2 annual growth is estimated at 6.2%, down from 6.3% in May, and June M1 annual change is estimated at a 0.9% contraction, down from a 0.8% contraction in May.
Updating the revision data from the prior Flash Update: annual M2 growth rates for March, April and May, respectively, were 6.1%, 6.5% and 6.7%, before revision, and 5.6%, 6.1% and 6.3%, after revision, with the preliminary June estimate at 6.2%. Annual SGS Ongoing M3 growth rates for March, April and May, respectively, were 11.7%, 12.8% and 13.3%, before the M2 and large time deposit revisions, and 11.5%, 12.5% and 13.1% after, with June coming in at a preliminary 13.0%. Estimates of the revised M3 monthly average levels will be published with the full-month data in the upcoming newsletter.
___________________________________________
The June/July "Shadow Government Statistics" monthly newsletter should be posted on July 16th. An e-mail advice of same will be sent.
FLASH UPDATE
July 11, 2007
__________
Revision and Seasonal-Adjustment Games Help Obfuscate Employment Reality
__________
The SGS Ongoing M3 Data Series has been updated for June's preliminary estimate, based on 25 days worth of data. The full-month estimate will be posted this coming weekend and discussed in the newsletter. The June/July newsletter is planned now for posting on July 16th, along with the Hyperinflation Summary. All postings will be advised by e-mail. -- Best wishes to all -- John Williams
To maintain the reported 2,000,000 (exact) annual jobs growth in place for revised seasonally-adjusted May 2007 payrolls, monthly jobs gain reporting needs to be targeted by the Bush Administration at 167,000 per month. In contrast, the Clinton Administration targeted 250,000 per month (3,000,000 per year) for an extended period of time. Perhaps for fear of rattling the credit markets, however, initial monthly reporting typically has been "understated" over the last year or so, followed in subsequent months by major upward revisions to prior reporting. If this pattern is not a machination of the Bureau of Labor Statistics (BLS), and if the later numbers are accurate, not fabrications, the BLS would do well to suspend its initial reporting of these numbers, rather than to continue misleading the public and the markets with such poor quality reporting.
Massive prior-period revisions and unusual seasonal-adjustments distorted the June employment report. Seasonally-adjusted June payrolls were reported up by 132,000 (207,000 net of revisions) +/- 129,000 (95% confidence interval), following a revised 190,000 (previously 157,000) gain in May. With the BLS re-jiggering its seasonal factors each month, the range of potentially gimmicked reporting is great. Nonetheless, year-to-year change in the seasonally-adjusted and unadjusted series should be the same, and based on June's unadjusted annual growth rate, the jobs increase should have been 107,000 instead of 132,000. Annual growth for June was reported at 1.5% seasonally adjusted, 1.4% unadjusted.
Adding to the malleability of the reporting, June's fudge or bias factor (birth/death model) created 156,000 jobs that were added to the unadjusted number for good measure. That number flows through to the seasonally-adjusted data with minimal variation. The bias factors never are adjusted for a recessionary environment, always assuming ongoing growth. For example, 25,000 construction jobs were added to employment last June, and 26,000 construction jobs were fudged this June, despite the obvious deterioration in the 2007 construction environment.
Further, the reported 16,500 June jobs gain in the educational sector suggests that, as usually is the case, the BLS had problems properly seasonally adjusting for the effect of the end of the school year.
The household survey counts those who are employed, or not, while the payroll survey counts the number of jobs. Even allowing for the reported 7.5 million individuals holding multiple jobs, the two series are irreconcilable by the BLS. The household survey showed seasonally-adjusted June employment up by 197,000, with seasonally-adjusted U.3 unemployment increasing to 4.53% +/- 0.2% in June, from 4.46% in May. Unadjusted U.3 rose to 4.7% in June from 4.3% in May.
The broader, seasonally-adjusted U.6 unemployment rate held at 8.2% in June, but rose on an unadjusted basis to 8.5% from 7.9% in May. Net of the "discouraged workers" defined out of existence during the Clinton Administration, traditional unemployment continues to run around 12%.
Other economic reporting last week was light, except for some continuing strength in the purchasing managers surveys, which will be discussed in detail in the upcoming newsletter.
Ongoing M3 Growth at 13% for June. Based on 25 days of reporting in June and the major revisions to small time deposits (M2) and large time deposits (M3) discussed in the July 1st Flash Update, year-to-year growth for June M3 is estimated at 13.0%, down from 13.1% in May. June M2 annual growth is estimated at 6.2%, down from 6.3% in May, and June M1 annual change is estimated at a 0.9% contraction, down from a 0.8% contraction in May.
Updating the revision data from the prior Flash Update: annual M2 growth rates for March, April and May, respectively, were 6.1%, 6.5% and 6.7%, before revision, and 5.6%, 6.1% and 6.3%, after revision, with the preliminary June estimate at 6.2%. Annual SGS Ongoing M3 growth rates for March, April and May, respectively, were 11.7%, 12.8% and 13.3%, before the M2 and large time deposit revisions, and 11.5%, 12.5% and 13.1% after, with June coming in at a preliminary 13.0%. Estimates of the revised M3 monthly average levels will be published with the full-month data in the upcoming newsletter.
The June/July "Shadow Government Statistics" monthly newsletter should be posted on July 16th. An e-mail advice of same will be sent.