There was also a massive, though preliminary, benchmark revision to hiring from April of 2005 to March of 2006. BLS is thinking about adding another 810k jobs to those that had been officially counted so far, after reviewing state employment tax records. That would be the biggest revision to a yearly job count since BLS began using state records to audit its payroll job tally, in 1991. Given that the work is preliminary, and that the revision is about 3 times larger than normal, it may turn out that the final revision, to be released next Febuary, is smaller than 810k. For now, BLS guys say they are trying to figure out how such a large discrepancy could have occurred.
Also interesting, I think, is that the Q3 wage gain was the slowest of any quarter this year. Bad for workers, since this is just about the last chance for a sizeble gain in take home pay in this cycle. Good for the Fed, though, as this is the time they fear that the cclical component of inflation can become, as they say, "entrenched" through wages. It seems the slowdown in hiring this year is cooling wages ahead of schedule.
Any ideas about the huge upward revision in the August numbers? And what typically causes these revisions?
If I were attempting to make a bullish case I would say that a similar revision to this months numbers would bring them in line with economists expectations and since the report is otherwise strong (except for the wages), there is really nothing to worry about.
What I find interesting is that we have a low total, but overall construction, and even real estate were positive. Commerical construction absorbed the decrease in residential, and then some.
Omark, total residential construction is from two BLS categories under construction:
1) Residential building
2) Residential specialty trade contractors
ac, these revisions bounce around (there is no pattern like with New Home Sales). However the benchmark revisions that k harris mentioned is interesting and might reveal a flaw in the BLS system.
Our host is right, it's noise. For what it's worth, the noise in the August revisions was largely in private services, which accounts for 43k of the revision. Trade and transport accounts for 12k. Health and social assistance accounts for 17k. Leisure and hospitality 19k. Those smaller components add to 48k, but there are other categories that were revised down instead of up. Those seem to be the big ones, though. The thing to note is that these are all categories showing pretty good growth this year, so the upward revisions don't change the impression of directon or pace of change for the categories. Same picture, different numbers.
My suspicions are that these peculiar numbers are indicative of something else - an economic "POP" of sorts. Knowing what little I do about these sort of economic events I would expect various economic indicators to accelerate and then suddenly drop off sharply. I also wouldn't expect these factors to be entirely in sync, so I would expect a bursting bubble economy to go through a period of contradictions where some indicators still show strong growth while others suddenly start falling.
It's just an ad hoc theory and I really haven't investigated past scenarios, but at least the "weird" report we had today seems to fit the theory.
As someone noted on CNBC this morning, the construction employment number does not include illegals - who tend to be fired before legal workers. One would expect that there is would be a larger drop in total employment if one were to include illegals.
Overall, the upwards revision to the August report offset the weak September report.
CR, this is a strong statement I'm not so sure I agree with. I don't buy their spin at all. The rule of thumb is that 150,000 jobs need to be created each month to accomodate new entrants in the job market. That is, 150,000 is just treading water. (51,000+188,000)/2=119,500. Not good enough, methinks.
You will get small little bursts here and there. September "may" have been the beginning of a new downward motion that sets even slower job growth. We shall see.
Overall, the upwards revision to the August report offset the weak September report - CR
CR, this is a strong statement I'm not so sure I agree with. - Emmanual
That's why he runs the long term running total on a chart with 'control limits' - so in one look a month up, a couple more down, a month flat and a month up can be understood in a glance without picking up the calculator even once.
Now, I'm no fan of this administration - I think they are about as sharp as a box of rocks, well worn round rocks - but so far since 2004 job growth hasn't been one of their problems. The chart with the upper & lower limits makes that much pretty clear - assuming you judge those limits to be reflective of good & bad performance.
Emmanuel - 150,000 as the standard to tread water has changed to around 100k-130k. I'm not sure how true it is.
"Because fewer people are entering the labor force than in years past, a smaller number of new jobs is needed to keep the unemployment rate constant, economists said. That number is now about 130,000 per month, Fed Chairman Ben S. Bernanke said in testimony to Congress in July, while Chicago Fed President Michael Moskow put the figure at 100,000. The number used to be as high as 150,000, economists say. "
LOL, dryfly & Steve. At this rate, by the end of Bush's term, they might say that zero new jobs per month is good enough because many are "leaving the labor force to join the life-of-leisure sector."
Good ol' lower labor participation rate--the black hole of employment.
Forchrisake wcc 'seasonal' means 'salted', not merely spring, summer or fall. That is so wintered.
I can do this straight: 810k/yr is a monumental admission of incompetence. Of course only if you ignore that the 12M illegal aliens is a guess...part of that National Intelligence Estimating.
Alright I can't do anything straight on a Friday.
This was a guess.
"There was also a massive, though preliminary, benchmark revision to hiring from April of 2005 to March of 2006. BLS is thinking about adding another 810k jobs to those that had been officially counted so far, after reviewing state employment tax records."
Does this mean that the average per hour wage is even lower than what was being reported?
good question EZ and I'd say yes after visiting the BLS site and finding this bit:
Each year, the Current Employment Statistics (CES) survey data are benchmarked to comprehensive counts of employment for the month of March derived from state unemployment insurance (UI) tax records that nearly all employers are required to file. For national CES series, the annual benchmark revisions over the last 10 years have averaged plus or minus two-tenths of one percent. The preliminary estimate of the benchmark revision for March 2006 is +810,000 (0.6 percent).
[Why do they hide the 0.2 percent by writing it out so that a visual comparison is avoided? I'm not cynical yet.]
It is a large number but not in terms of the total employed,(~144M) so average wages will not be seriously overstated. But check carpenters out at ~$8/hr and know that this pitiful number is likely overstated and not, say, the healthcare workers. But they are investigating the sources of this possible restatement...sure, my cynicism is under control, yours?
"It is a large number but not in terms of the total employed,(~144M) so average wages will not be seriously overstated. But check carpenters out at ~$8/hr and know that this pitiful number is likely overstated and not, say, the healthcare workers. But they are investigating the sources of this possible restatement...sure, my cynicism is under control, yours?"
Yes it is small compared to the total employed. I don't have details yet, but the 810 k jobs divided by 12 months is 67,500 jobs per month. Add that to the jobs number each month. The employment numbers for the period look exceptional well.
There was also a massive, though preliminary, benchmark revision to hiring from April of 2005 to March of 2006. BLS is thinking about adding another 810k jobs to those that had been officially counted so far, after reviewing state employment tax records. That would be the biggest revision to a yearly job count since BLS began using state records to audit its payroll job tally, in 1991. Given that the work is preliminary, and that the revision is about 3 times larger than normal, it may turn out that the final revision, to be released next Febuary, is smaller than 810k. For now, BLS guys say they are trying to figure out how such a large discrepancy could have occurred.
Also interesting, I think, is that the Q3 wage gain was the slowest of any quarter this year. Bad for workers, since this is just about the last chance for a sizeble gain in take home pay in this cycle. Good for the Fed, though, as this is the time they fear that the cclical component of inflation can become, as they say, "entrenched" through wages. It seems the slowdown in hiring this year is cooling wages ahead of schedule.
Any ideas about the huge upward revision in the August numbers? And what typically causes these revisions?
If I were attempting to make a bullish case I would say that a similar revision to this months numbers would bring them in line with economists expectations and since the report is otherwise strong (except for the wages), there is really nothing to worry about.
What I find interesting is that we have a low total, but overall construction, and even real estate were positive. Commerical construction absorbed the decrease in residential, and then some.
I'm a little confused about the residential construction number -- what items are you including?
Omark
Omark, total residential construction is from two BLS categories under construction:
1) Residential building
2) Residential specialty trade contractors
ac, these revisions bounce around (there is no pattern like with New Home Sales). However the benchmark revisions that k harris mentioned is interesting and might reveal a flaw in the BLS system.
Best to all.
hey -- didn't I read this morning that the consensus prediction for the September number was 125,000? That seems like a big difference, wonder why?
AC,
Our host is right, it's noise. For what it's worth, the noise in the August revisions was largely in private services, which accounts for 43k of the revision. Trade and transport accounts for 12k. Health and social assistance accounts for 17k. Leisure and hospitality 19k. Those smaller components add to 48k, but there are other categories that were revised down instead of up. Those seem to be the big ones, though. The thing to note is that these are all categories showing pretty good growth this year, so the upward revisions don't change the impression of directon or pace of change for the categories. Same picture, different numbers.
harris,
My suspicions are that these peculiar numbers are indicative of something else - an economic "POP" of sorts. Knowing what little I do about these sort of economic events I would expect various economic indicators to accelerate and then suddenly drop off sharply. I also wouldn't expect these factors to be entirely in sync, so I would expect a bursting bubble economy to go through a period of contradictions where some indicators still show strong growth while others suddenly start falling.
It's just an ad hoc theory and I really haven't investigated past scenarios, but at least the "weird" report we had today seems to fit the theory.
As someone noted on CNBC this morning, the construction employment number does not include illegals - who tend to be fired before legal workers. One would expect that there is would be a larger drop in total employment if one were to include illegals.
Overall, the upwards revision to the August report offset the weak September report.
CR, this is a strong statement I'm not so sure I agree with. I don't buy their spin at all. The rule of thumb is that 150,000 jobs need to be created each month to accomodate new entrants in the job market. That is, 150,000 is just treading water. (51,000+188,000)/2=119,500. Not good enough, methinks.
You will get small little bursts here and there. September "may" have been the beginning of a new downward motion that sets even slower job growth. We shall see.
Overall, the upwards revision to the August report offset the weak September report - CR
CR, this is a strong statement I'm not so sure I agree with. - Emmanual
That's why he runs the long term running total on a chart with 'control limits' - so in one look a month up, a couple more down, a month flat and a month up can be understood in a glance without picking up the calculator even once.
Now, I'm no fan of this administration - I think they are about as sharp as a box of rocks, well worn round rocks - but so far since 2004 job growth hasn't been one of their problems. The chart with the upper & lower limits makes that much pretty clear - assuming you judge those limits to be reflective of good & bad performance.
That may all change but so far so good.
Emmanuel - 150,000 as the standard to tread water has changed to around 100k-130k. I'm not sure how true it is.
"Because fewer people are entering the labor force than in years past, a smaller number of new jobs is needed to keep the unemployment rate constant, economists said. That number is now about 130,000 per month, Fed Chairman Ben S. Bernanke said in testimony to Congress in July, while Chicago Fed President Michael Moskow put the figure at 100,000. The number used to be as high as 150,000, economists say. "
LOL, dryfly & Steve. At this rate, by the end of Bush's term, they might say that zero new jobs per month is good enough because many are "leaving the labor force to join the life-of-leisure sector."
Good ol' lower labor participation rate--the black hole of employment.
The retail YoY number is -0.6%, not seasonally adjusted. Why use seasonally-adjusted numbers when YoY alone should get rid of any seasonality?
All those purty pictures illustrate a still very strong economy.
Retail weak? Nope.
Strong increase in retail market - Chicago Tribune
Don't let that stop you ladies from the hand wringing, though.
Yes, Retail is weak, and no the economy has not been strong in Q3, look at those falling durables.
Forchrisake wcc 'seasonal' means 'salted', not merely spring, summer or fall. That is so wintered.
I can do this straight: 810k/yr is a monumental admission of incompetence. Of course only if you ignore that the 12M illegal aliens is a guess...part of that National Intelligence Estimating.
Alright I can't do anything straight on a Friday.
This was a guess.
Thanks for posting this as the day job had me snowed today. I just posted a wee bit more including the usual garbage from our President.
"There was also a massive, though preliminary, benchmark revision to hiring from April of 2005 to March of 2006. BLS is thinking about adding another 810k jobs to those that had been officially counted so far, after reviewing state employment tax records."
Does this mean that the average per hour wage is even lower than what was being reported?
good question EZ and I'd say yes after visiting the BLS site and finding this bit:
Each year, the Current Employment Statistics (CES) survey data are benchmarked to comprehensive counts of employment for the month of March derived from state unemployment insurance (UI) tax records that nearly all employers are required to file. For national CES series, the annual benchmark revisions over the last 10 years have averaged plus or minus two-tenths of one percent. The preliminary estimate of the benchmark revision for March 2006 is +810,000 (0.6 percent).
[Why do they hide the 0.2 percent by writing it out so that a visual comparison is avoided? I'm not cynical yet.]
It is a large number but not in terms of the total employed,(~144M) so average wages will not be seriously overstated. But check carpenters out at ~$8/hr and know that this pitiful number is likely overstated and not, say, the healthcare workers. But they are investigating the sources of this possible restatement...sure, my cynicism is under control, yours?
"It is a large number but not in terms of the total employed,(~144M) so average wages will not be seriously overstated. But check carpenters out at ~$8/hr and know that this pitiful number is likely overstated and not, say, the healthcare workers. But they are investigating the sources of this possible restatement...sure, my cynicism is under control, yours?"
Yes it is small compared to the total employed. I don't have details yet, but the 810 k jobs divided by 12 months is 67,500 jobs per month. Add that to the jobs number each month. The employment numbers for the period look exceptional well.