Summary
On Friday, February 3rd, the Labor Department released January employment data, indicating that the unemployment rate fell 0.2% during the month, to 4.7%. Also reported was a 193,000 gain in payroll jobs. The former number was better than expected; the latter was worse. Meanwhile -- and again, wage growth did not keep pace with the Consumer Price Index.
_____
NOTE: The recent employment report included revisions to both the household and establishment surveys, per the following announcement:
"The establishment survey data in this release have been revised as a result of the annual benchmarking process and the updating of seasonal adjustment factors... In addition, household survey data for January 2006 reflect updated population controls... Also, new seasonally adjusted employment data for multiple jobholders have been added..."
All past data discussed or shown in this missive reflect the above revisions unless otherwise noted. In addition, appearing later in this missive is John Williams' critical analysis of some of what the Labor Department did in formulating these revisions. Fasten your seatbelt!
* February 3rd's release from the Labor Department indicated that the unemployment fell 0.2% during January, to 4.7%. The report also showed an increase of 193,000 payroll jobs during the month.
* The consensus forecast was looking for an unemployment rate that would remain unchanged at 4.9%, versus the 4.7% that was reported. The consensus prediction for the gain in payroll jobs was around 250,000, well above the actual number reported. However, net revisions to November and December data more than offset this shortfall.
* The two components comprising each month's employment report are the household survey and the establishment survey (the latter also known as the "payroll survey.") The unemployment rate is derived from the former, making it a visible, highly sensitive number for political purposes.
* The payroll survey represents the number generally more important to the immediate behavior of the financial markets, although it also possesses a good deal of political importance. The recent report indicated job growth of 193,000 during January. This compared with a gain of 76,000 in payroll employment during January of last year.
* February 3rd's report contained net revisions that added 81,000 to November and December payroll job creation. November went to +354,000 from +305,000, while December went to +140,000 from +108,000. Thus, including revisions, payroll employment over the last three months grew by 687,000 jobs, versus the 606,000 gain before the November and December adjustments. (During the same three-month period in 2004-2005, payroll employment rose by 369,000.)
The following table breaks out the reported growth in payroll employment over the trailing three months, versus the same period in each of the prior five years.
------------------------------------------
TRAILING THREE-MONTH PAYROLL EMPLOY-
MENT GROWTH (Thousands of Workers)
------------------------------------------
2006/ 2005/ 2004/ 2003/ 2002/ 2001/
Month 2005 2004 2003 2002 2001 2000
------------------------------------------
Jan. 193 76 122 82 -113 -13
Dec. 140 160 122 -121 -162 158
Nov. 354 133 79 -16 -308 216
----------------------------------
Total 687 369 323 -55 -583 361
=== === === === === ===
------------------------------------------
NOTE: After revision, payroll employment expanded by 1.976 million jobs during the 12-month period ended December, 43,000 less than the 2.019 million prior to revision. However, the pattern of the revisions incorporated by the Labor Department tended to reduce growth during earlier months, while increasing it during later ones. For instance, after revision, payroll job creation was reported at 0.754 million during the five months ended December, versus 0.603 million prior to revision.
* For the 12 months ended January 2006, payroll employment expanded by a reported 2.093 million jobs. This compared with an increase of 2.051 million during the 12-month period ended January 2005. Over the six-month periods ended January 2006 and 2005, reported growth was 0.947 million and 1.015 million, respectively.
NOTE: The report also contained benchmark revisions to prior earnings data. These are incorporated in the following discussion and in the table that follows.
* According to February 3rd's report, average hourly earnings rose seven cents during January, or by about 0.4%, to $16.41. This compared with December's revised $16.34.
* Year over year, average hourly earnings rose 3.3%. This compared with a Consumer Price Index that rose 3.4% for the 12 months ended December (the latest CPI data currently available).
* During the 12 months ended December, the year-over-year gain in the CPI exceeded the year-over-year growth in average hourly earnings in all but one of those months (June 2005), exhibiting the longer-run loss of real purchasing power based on these particular series. And that is against a CPI that we (and the vast majority of others, too) believe is significantly understated.
* The table below contains the latest available results for key employment and inflation series. (NOTE: The new material contained in these tables will be posted to the "Data & Charts" section on the GRA website.)
-------------------------------------------------
KEY U.S. EMPLOYMENT AND INFLATION DATA
-------------------------------------------------
PAYROLL
UNEMPLOYMENT EMPLOYMENT AVERAGE HOURLY
RATE CHANGE EARNINGS
(%) (000s) ($s)
--------------- ---------- -----------------
Month Yr. Yr. Yr. Year/
/Year No. Ago No. Ago No. Ago Year
--------------- ---------- ------------------
01/06 4.7 5.2 193 76 16.41 15.88 3.3%
=================================================
12/05 4.9 5.4 140 160 16.34 15.84 3.2%
11/05 5.0 5.4 354 133 16.28 15.81 3.0%
10/05 4.9 5.4 37 338 16.28 15.79 3.1%
09/05 5.1 5.4 48 177 16.19 15.77 2.7%
08/05 4.9 5.4 175 131 16.16 15.74 2.7%
07/05 5.0 5.5 241 38 16.14 15.68 2.9%
06/05 5.0 5.6 166 78 16.07 15.65 2.7%
05/05 5.1 5.6 106 267 16.03 15.63 2.6%
04/05 5.1 5.5 228 252 16.00 15.58 2.7%
03/05 5.1 5.7 140 346 15.95 15.54 2.6%
02/05 5.4 5.6 265 55 15.91 15.52 2.5%
-------------------------------------------------
EMPLOY. COST INDEX (%)* PRODUCTIVITY (%)#
----------------------- ------------------------
Year Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Yr.
----------------------- ------------------------
2005 0.7 0.7 0.8 0.8 3.2 2.1 4.5 -0.6 2.7
2004 1.1 0.9 0.9 0.8 2.1 4.5 1.3 2.5 3.4
2003 1.2 0.9 1.0 0.8 3.1 6.6 9.6 0.8 3.8
2002 0.8 1.0 0.8 0.8 6.5 0.8 4.1 0.2 4.0
2001 1.1 1.0 1.0 1.0 -0.4 5.6 1.5 6.5 2.5
2000 1.2 1.0 0.9 0.9 -2.0 7.2 -0.9 4.0 2.7
1999 0.4 1.1 0.9 1.1 3.6 0.5 2.8 7.1 2.8
1998 0.7 0.8 1.0 0.7 3.2 1.2 4.5 2.1 2.8
1997 0.7 0.8 0.8 1.0 -1.2 5.0 3.4 1.6 1.6
-------------------------------------------------
CONSUMER PRICE INDEX## PRODUCER PRICE INDEX@
-------------------------- ---------------------
Month Yr. Year/ Yr. Year/
/Year Index Ago Year Index Ago Year
-------------------------- --------------------
12/05 197.7 191.2 3.4% 160.3 151.7 3.6%
11/05 197.8 191.2 3.5% 158.9 152.1 4.5%
10/05 198.9 190.7 4.3% 160.0 151.1 5.9%
09/05 198.5 189.6 4.7% 158.9 148.9 6.7%
08/05 196.1 189.3 3.6% 156.2 148.5 5.2%
07/05 195.1 189.2 3.1% 155.2 148.3 4.7%
06/05 194.1 189.3 2.5% 153.7 148.2 3.6%
05/05 194.1 188.8 2.8% 153.7 148.3 3.6%
04/05 194.2 187.7 3.5% 154.4 147.7 4.7%
03/05 193.2 187.3 3.2% 153.7 146.4 5.0%
02/05 192.0 186.5 2.9% 152.5 145.6 4.7%
01/05 191.3 185.9 2.9% 151.9 145.7 4.3%
-------------------------------------------------
NOTE: All data are seasonally adjusted. *Total
civilian compensation, 3-month change. #Nonfarm
business output per hour, annual rates. ##All
urban, all items. @Finished goods.
-------------------------------------------------
Some Thoughts from John Williams
NOTE: Because we are now at the time of year when most government economic series will undergo so-called "benchmark" revision, I thought clients would find valuable knowing something about the bizarre machinations that have found their way into the process, using employment data as the proxy. Thus, the following observations from colleague, John Williams.
John is genuinely expert in analyzing this nonsense, "nonsense" a euphemism for something less forthright from the folks in Washington. If you are not familiar with John's work, be sure to visit his website, Shadow Government Statistics.
_____
"As supposedly finalized last month, the BLS expected to revise unadjusted payroll levels for March 2005 downward by 203,000. The amount ended up to be a downward revision of 158,000, but its application was far from uniform.
"The pattern of actual revisions shows a bowl shape that results from the pattern of applying the revisions to the unadjusted data combined with shifts in seasonal adjustment patterns. Somehow, the BLS's reporting problem largely corrected itself over time. As a result, the effects of the benchmark were to minimize the impact on current reporting, leaving intact the already reported two million jobs 'creation' of the last year that has been so heavily touted by the administration.
"Changing population assumptions, as was done in January, is a normal part of estimating unemployment as measured in the household survey. Yet, January's improved unemployment rate cannot be compared to December's reading, because the data were inconsistent. There was a time when the BLS published data both ways so comparisons could be made, but the lack of same, now, does not seem to bother anyone in the financial media.
"Thanks to some unusual benchmark revision patterns, January payrolls rose by more than 100,000, but remained shy of market expectations. As usual, the BLS was unable to seasonally adjust successfully the unemployment data as the workforce transitioned from the holiday to the post-holiday season, and the reported unemployment rate fell. Separately, the household data included updated 'population controls,' which makes the comparisons of inconsistent December and January data meaningless.
"The labor data remain largely worthless, as the payroll revisions skewed reporting and did nothing to resolve the obvious reporting problems still surrounding the effects of Hurricane Katrina. The household revisions made the data incomparable, again, while showing that the seasonal adjustments remain consistent distorting factors for the reported data.
"The popularly followed seasonally-adjusted unemployment rate U-3 for January fell to 4.69%, down from December's 4.91%, a change just outside the published +/- 0.2% error margin, but the data are inconsistent and properly should not be compared. Unadjusted U-3 unemployment rose to 5.1% in January, from December's 4.6%, while the broader U-6 unemployment measure jumped to 9.2% in January from December's 8.4%. January's seasonally-adjusted U-6 rate, however, eased to 8.4% against December's 8.6%. Including the long-term 'discouraged workers' defined away during the Clinton administration, total unemployment remains roughly 12 percent.
"Because of the break in the series between December and January, the BLS did not even try to report the monthly change in household employment.
For January, the household survey showed a seasonally adjusted gain of 193,000 (96,000 net of revisions) +/- 108,000, following December's 140,000 gain (was 108,000). Annual growth in unadjusted January payrolls was 1.57%, up from December's 1.48%.
"January's payroll gain of 193,000 included a negative bias of 193,000 jobs from the 'net birth/death' adjustment to potential reporting entities, a level reduced from the prior year's downside bias of 280,000.
"The latest report was against a background of flat December help-wanted advertising, negative employment trends as reported in the various January purchasing managers surveys, and improved new claims for unemployment insurance."
_____
|
