A little weaker than I would have expected from ISM and factory orders. Its not that mfg is that large a factor anymore by itself but it is a 'driver' of activity in other sectors... mfg activity create or support lots of other jobs in services, transportation & such.
I would say the ISM survey might be bogus but not BOTH orders & survey. Maybe there is a lag.
The report looks even worse if one focuses on the household survey. In my Angrybear tradition, I graph the employment to population ratio - which alas dropped to 63.0%.
I am still not buying this talk of hundreds of thousands of job losses in residential construction. The jobs might appear to go but in reality many of the workers do find alternate work either in a more casual capacity by word of mouth or by doing something different where currently it is hard to find workers. The overall effect is therefore a softening rather than an over the cliff type of response.
Manufacturing employment declined, and was actually stronger than it should have been due to a strike that ended (accounting for 6,500 jobs). Finance and insurance lost 14,000.
Retail was down 26,000 net.
Did you see the results of the household survey? According to Marketwatch: "The separate survey of households showed employment fell sharply by 468,000 jobs in April. This is the biggest drop since November 2002. Unemployment rose by 77,000 to 6.80 million. The civilian labor force declined by 392,000." That is a sizeable drop, and a pretty big divergence from the establishment survey.
I remember at the outset of the recovery the household survey consistently "beat" the establishment survey in job creation. At the time, this was seen as evidence that the recovery was driven by a burst of entrepeneurial activity.
Could it be that, in reality, the household survey was better at counting residential construction workers? And that now we are seeing the reverse?
The household survey stat will certainly be ignored by the bulls, but even this payroll one is heavily revised, no?
The professional services contribute the most, yet the weekly wage actually moves down. Sounds like less hours to me.
Last nagging thing: the wait for construction jobs to disappear might just take longer or we face the music that the work performed by these "material movers" (as the BLS describes it) was not counted when it was performed and therefore not counted when it shifted to agricultural labor where it is seasonal, temporary and generally not "payrollable". There is a certain asymmetry about this undocumented insourcing and under/misdocumented outsourcing that keeps productivity numbers high.
Jed - Doesn't the household survey have much larger uncertainty? Order of magnitude with the actual 'signal'? I know the bulls usually trumpet household over establishment (at least the last few years as household was considerably stronger)... and bears usually complain about the error bars... isn't that the same problem now (but in reverse)?
I'd love to hear/read more detail on this... and not just gov't boiler plate... someone who could EXPLAIN the gov't boiler plate.
Is military personnel included in the government employment category. If so is the war contributing to the strength in the government employment numbers?
The business birth/death plug in the April payroll number was +317k, which is the largest ever for this statistical invention. I would think with a slowing economy this plug will prove quite optimistic (the April plug has almost always been revised down anyway) and the headline number will be revised down close to zero.
I've been more in the long period of very weak growth camp, but, in my mind, business confidence picking up (ism surveys, m&a activity) with the consumer coming under increasing pressure makes a hard landing more likely. This is because business is ramping up at precisely the wrong time, and recessions are generally caused by businesses scaling back inventories and capital investment (consumer spending rarely goes negative for any length of time).
calmo - you are 100% correct about productivity being overestimated due to poor man-hour accounting. I've talked with mfg wonks and they would agree strongly with your comment.
This is because business is ramping up at precisely the wrong time, and recessions are generally caused by businesses scaling back inventories and capital investment (consumer spending rarely goes negative for any length of time).
Turbo | 05.04.07 - 10:23 am | #
Could be but understand that companies don't 'ramp up' inventories very long anymore... couple very limited inventory space with tightly coupled supply chain JIT systems & there isn't anywhere to put the stuff. The warehouse is also the retail shelf.
If they have initially missed the signal - they won't miss it for long... the stuff will pile up in the isles if nothing else.
Watch consumption - that is the last link in the supply chain that stretches from the raw ore in the mine to the crap in your garage. Consumption is the last valve in the pipeline.
The lender didn't have a license, misapplied payments, claimed she didn't make them and failed to pay her taxes out of escrow, leading to a mistaken foreclosure filing.
The expected job losses in residential construction haven't happened because the idiot builders are still building despite a gargantuan inventory of unsold houses.
Like so much of what is happening at the moment, this will delay the crash somewhat. Right now the rich, the powerful and the connected are frantically pulling every lever of power and propaganda in the hope of getting one or two more quarters of big bonuses.
But the crash is coming. And the longer it is delayed and denied the worse it will be when it inevitably descends upon us.
Dryfly - I agree. That's why the business cycle has become, well, less cyclical and that makes it hard for the gdp math to go negative (even for a bear like me) unless the consumer really craters. I guess the point I was trying (poorly) to make it that the current uptick in business confidence is unlikely to hold up for too long against a slowing consumer, and unlike the ravers trading stocks right now, I'd put much more weight on consumer measures, rather than business measures, as you suggest.
That being said, even though it is likely to be short-lived, a ramp-up in business activity now is just another overhang weighing down the economy by this summer. If we can go a week without another large buyout being announced, and the stock market stops going vertical, then the financial system will have to start dealing with its considerable problems, which will start the ball rolling in a way that makes me very nervous right now.
"This is probably just the beginning of the loss of hundreds of thousands of residential construction jobs over the next year or so."
Never mind that you've done little more than meticulously back up this projection with comprehensive statistical analysis -- Your conclusion remains utterly irresponsible!
(Just trying to provide a public service here ...)
I recall that the red line in earlier iterations of the top graph were more or less dead center in the range of expected outcomes. This month, the red line has come back into the expected range, after an excursion into happy territory. Am I remembering rightly? If so, what has changed about the graph?
Dryfly,
I wouldn't put too much emphasis on one month's headline factory orders data for guidance on hiring. I don't think the correlation is all that high. Moreover, factory orders are down 1.3% y/y, and orders ex-transport have only risen in 3 of the past 9 months. The factory hiring diffusion index is at its lowest level since late 2003. If I were doing the hiring, I wouldn't look for more workers under those conditions.
Steve,
Quite right that ADP has improved in its new version. I would note, though, that ADP anticipated fewer factory job losses in April, and that isn't what happened. Still, it's good to see improvement.
I wonder what employment statistics are for the real estate industry in and of itself. How many agents can keep hanging on when they have to share a smaller and smaller pool of buyers and sellers? At potentially lower prices?
I use the establishment survey because the data is "harder", less volatile and has a dramatically larger sample-size than the household data. I also use it consistently and exclusively over time, regardless of whether it supports my position or not.
The jobs might appear to go but in reality many of the workers do find alternate work either in a more casual capacity by word of mouth or by doing something different where currently it is hard to find workers.
Worried, that's the textbook definition of under-employment. It has the same depressive effect on economic activity.
Sebastian, you have your work cut out for you today. But don't despair, you have plenty of like minded folk in the captive financial media. The bad job numbers simply mean the Fed has less incentive to raise interest rates. Of course, that also poses a problem for people who care about a weakening dollar. But then again, a weak dollar is a good thing, too, because that means U.S. internationals earn more $ on foreign exchange.
In the meantime, however, things aren't looking so rosy for those of us on Main Street as we watch our jobs go away and our purchasing power evaporate.
Quite right that ADP has improved in its new version. I would note, though, that ADP anticipated fewer factory job losses in April, and that isn't what happened. Still, it's good to see improvement.
Yeah, some of the internals didn't quite match and they predicted large downward revisions for March, so their March number is even more off. Still, it looks like the recent changes improved things a bit.
Sebastien - Intimidating BLS it isn't. Incomplete it is.
I own/run a small biz - mostly family members dropping in & out. The wife & I fill out the establishment survey every quarter so I know what they are asking - it can't tell the whole story 'cause they don't ask the right questions.
In our case revenue is 1099 from clients to my company & then either salary or 1099 to us individually. We are not Sub S but run like a Sub S so it is a straight pass through & avoids double taxation. We pay the taxes individually like a partnership.
Most of what we 'do' doesn't show up immediately in establishment - I do but that's about it, rest of the 'work' folks do for me I 'contract'.
And believe me they are free agents & part-time... when you work with family cats start looking like reliable alternatives.
Many of the people I know are in the same boat. Even some jobs that used to be 'salaried' are now contract... For example I am having discussions with a couple companies now where I'll be their 'New Business Dev Director' - basically a hired gun hunter. Won't ever show up in their establishment anywhere... it will look like a 'job loss' on paper but isn't.
My daughter might work with me on this (she's a double engineer - marketing major)... she won't show up anywhere period either UNLESS I decide to 'hire her'... say to try & get her on a better insurance package under my company. If that happens then she'll show up on my establishment - probably not for many months afterward.
Eventually - most all eventually get captured in establishment somewhere but it is a real slow process. And they are just as likely to not be working for by then!
Like I said I know MANY people doing this - companies benefit by getting us off their benefits & pension roles & out of their offices... We benefit by not having bosses ride our ass about how & when we do what we do.
Of course getting paid is sometimes an issue - but that's 'off topic' here.
This trend is especially common in PEs & start-ups where the whole damned office might be 'equity holding' 1099 indy contractors. Lord knows there are a lot of PEs out there now.
But the BLS numbers don't accurately reflect this growing pool of labor - household tries to capture it but I don't think it does a great job - and this pool is VERY volatile, believe me, so it will never be an easy thing to do anyway.
BTW - I could just as easily see the trend to 1099 reverse IF pension & health care were no longer company responsibility... Execs hate working with indies like me, I'm also a bit like a cat that way... If I'm not making money then I'm blogging, or fishing, or watching grass grow... I don't do meaningless meetings nor cubicle time. Haven't for 20 years now.
Windshield time getting to and from clients & customers... or smokestack chasing... still doing lotsa that. Cubicles, no.
Oh, Dryfly, I couldn't agree more. This is one of the reasons why I still feel that this economy has a lot of longterm strength in it. Your account also fits a lot of banking, info, etc.
I think this is where the "business and professional services" comes in. It drives me nuts that I see no direct way to measure it, but this very real, very efficient, supports productivity, and, I am guessing, will cushion some of the negative impact from our credit excesses of recent years.
The expected job losses in residential construction employment still haven't happened"
A HB site in 1989 had lot more workers then the same site in 2005. I have wondered if the number of construction workers hired for the HB sector has been significantly overstated during this build up. BLS and others using outdated labor rates and not taking into account new technology and prefab aspects of the HB trade today.
Today NFP report suggests that the next recession can start 6 weeks from now:
What time?
Steve, monday morning, as usual - I can't say more accurately.
Seriously, a lot of stuff is pointing that recession can start as early as July.
But NFP was the last shoe to drop. Before today, payroll data suggested that recession is impossible in the next 2-3 months. But today datapoint suggests that July is possible.
Let me stress that I do not conclude recession from NFP. I conclude from everything else, but NFP is sort of show-stopper, which sais, for example, that recession is almost impossible, say, right now in May.
dryfly, thanks for the response. I completely agree with you that economic data is going to be incomplete and have flaws (as does virtually all economic data). The subsequent revisions make things even harder.
Yet it's the best we have, and I've found it to work well...as long as I'm aware of its limitations, what it is and isn't telling me. It beats guessing and it sure as Hell beats massaging the data to get a pre-determined answer.
"...Like I said I know MANY people doing this - companies benefit by getting us off their benefits & pension roles & out of their offices... We benefit by not having bosses ride our ass about how & when we do what we do..."
So, you're saying MANY people you know had their jobs outsourced, but those jobs stayed right here in the U.S.?
The government numbers were fudged upward by the largest amount yet in the birth/death fake job adjustment. They 'added' 317000 jobs that actually don't exist except in the BLS statistics department.
Don't believe me? Look at the link and study the methodology. It is the ultimate government fudge factor. Skew any bad data into good data.
Don't believe the government's numbers, they are lies.
The real fact is that 88000(reported)-317000(fictitious)=229000 jobs were lost last month. Yes, you read that right, the real numbers were a LOSS of 229000 jobs in April.
To continue somewhat with dryfly's description of white collar work that is not captured in the current employment surveys, I know several people including me in the blue-collar end that punch a clock for part of our income and then do side jobs to make ends meet.
Now the plural of anecdote is not data for our data-driven host but it does make me wonder if the current economy has changed enough to make revisions to the employment survey data collection methods necessary to accurately describe the current situation.
It seems to me our economy, despite the rabid growth in documentation for the simplest task, is much more informal than in past decades. And that informality may ignore the formal data collection process.
Jed wrote: "The separate survey of households showed employment fell sharply by 468,000 jobs in April. This is the biggest drop since November 2002. Unemployment rose by 77,000 to 6.80 million. The civilian labor force declined by 392,000." That is a sizeable drop, and a pretty big divergence from the establishment survey.
Exactly, see the birth/death adjustment lies to understand the divergence, and why our present government needs to be eliminated.
pgl, nice graph, and I hope policymakers are paying attention to the participation rate.
David Pearson, I prefer the establishment survey for total employment, but the household survey is very good for unemployment (and participation rate as pgl noted). The slight increase in the unemployment rate is insignificant - but the participation rate is important.
barely, don't forget Rafael Correa, Michelle Bachelet, Daniel Ortega, Evo Morales, Hu Jintao, Segolene Royal, Ándres Manuel López Obrador, and a hell of a lot more people in the USA than you would obviously like to believe.
Ordinarily dotty, I'd tell you that the birth/death model(s) have worked well over the past few years as evidenced by smaller subsequent revisions, but today I'm feeling less striped and more dotty...and I think their methods are not tested for significant trend changes and you could be spot on...but maybe only this time.
Calculated Risk, I want to suggest this answer to the conundrum.
The data are cloudy. But based on arguments I have made on my blog, I think that we are seeing real job growth in health care, probably low wage. As for manufacturing, there's too much internal inconsistency between the establishment survey and ISM to believe that factory performance is actually improving. Therefore, my guess is that factory orders did not significantly rise and that next month this will be evident.
The reason this is relevant is that health care spending on care for the elderly is another major economic problem the US faces. If employment for hospice care rises, it comes out either Medicare or savings: more debt on the consumer or on the government and therefore lower future growth. Whereas rises in civilian factory orders would probably imply an improvement in the trade situation.
There, I think, is the bull v. bear case in a nutshell. And so it all depends on where jobs were actually created, on the factory floor or in home health work.
Job quality is also something that isn't tracked with job creation figures: full time or part time? Contract, temp, or perm? Low-wage or high wage? Benefits, or no benefits?
Roll in all that and you might have the bare beginning of a meaningful metric as to the health of the job market. The ones we have are so much marsh gas.
dotcommunist said: "I'm confused, Steve and the other permabulls keep repeating the claim about that strong income growth, but somehow, I just keep seeing the opposite..."
While I agree that the birth-death model is a suspicious black box, the 317k is NOT seasonally adjusted and the not seasonally adjusted employment increased by 833k. So without B/D model you have 516k new jobs in April. I have no idea how to apply seasonal adjustment (I believe the seasonal adjustment is applied separately to each sector and the B/D model is also applied separately) to this figure but you cannot simply subtract 317k from 88k.
TJ, from the bls birth/death model page,http://www.bls.gov/web/cesbd.htm
"The most significant potential drawback to this or any model-based approach is that time series modeling assumes a predictable continuation of historical patterns and relationships and therefore is likely to have some difficulty producing reliable estimates at economic turning points or during periods when there are sudden changes in trend."
Basically, there's little way to figure out what these figures mean for the overall economy.
Lower employment figures MIGHT be a positive - reduces inflationary pressures. They also could be a negative, in that the economy is slowing.
But I usually ignore employment figures when trying to figure out where the economy is going for two reasons.
1- It's mostly a backward looking statistic.
2- If you go to the museum on Ellis
Island there's a great graph showing immigration over the past 200 years or so. There's a very marked pattern of reduced immigration and even emigration during "temporary downward periods of economic activity" (AKA recessions, depressions and panics). My grandmother, for example, was born in the US but moved back to Poland with her family during the recession in the early 1920s and then back to the US in the late 20s boom.
There's been a historically large amount of immigration over the past decade, or at least there has been in my neighbourhood - Manhattan is close to its highest ever levels of residents who were born overseas. Ive always suspected variance in immigration/emigration is important and isn't captured in the stats. I was very glad to see a similar point being made in a NY Times article last week.
More to point for this blog, the birth-death model added in 49,000 (not seasonally adjusted) construction workers in April. That's more than a fourth of the 188k reported in the release. It seems that in prior years, April was a pretty good time of year to start a residential contracting business. But the birth-death model has no way of knowing that those business births didn't happen this year, and in fact, there were probably quite a few mom-pop operations that decided not to build any homes this year.
The birth-death model is not a sham, nor are they "fictitious" jobs. But they are estimated jobs that will either be proven to have existed or be proven not to have existed once the hard numbers come in from the data like tax returns.
Most likely, when it's all said and done, we'll see construction jobs fell by at least 50,000 in April.
The bottom line: the report was very weak EVEN WITHOUT a huge decline in construction jobs that will likely show up in the revisions.
Dryfly, when I was contracting as an engineer years ago I noticed a pretty tight coupling between the direction of the prevailing economic winds and the availability and duration of contract jobs. They were usually the first cut when things started edging south. Have you seen any such slowdown out there?
The idea that BLS is somehow distorting the numbers is rather wild to me. I don't think the American taxpayer wants to pay for BLS surveying the entire population of households each month.
They do the best they can, and then they correct with additional data. The numbers are not going to be perfect.
Dryfly, when I was contracting as an engineer years ago I noticed a pretty tight coupling between the direction of the prevailing economic winds and the availability and duration of contract jobs. They were usually the first cut when things started edging south. Have you seen any such slowdown out there?
Andrew - I am the most counter-cyclic contract worker in America... I do better & am in more demand in busts than in booms.
My job is to find 'new business' for mfg companies and when they are full of work & have backlogs growing the last thing they want is 'new business' Then they tell me to go away. Not all that politely either.
Over the last year though I've been contacted more regularly but surprisingly not as much by US firms. Many of them have been consolidated into large PE takeovers and they don't need indy hired guns.
When the PEs buy a new firm it has a Sales & BD staff already in place so cherry picks the staff of the new firm for who they want then toss the rest out - they sure don't need more.
But offshore suppliers to US firms are really stepping up their presence here. They don't just want to work for the US owned FDI operations in Asia - they want to do what the Japanese did - go fully global, here as well as there.
People like me already know the lay of the land so they contact us - I listen to them, believe me.
Three of the four best new prospects I have are from outside the US... two from India, one from Canada... only one was a US firm (a brand new start-up).
I am reluctant to work with offshore accounts - too hard to make them pay you if they decide they aren't really interested... And who wants to sue them in Bangelore?
But that might be what I have to do if the US firms continue to consolidate like they have over the last couple years. If you can't join'em I guess you fight'em.
Thank God for the government, they added 25 of the 88 new jobs.
Overall this is a somewhat weak report.
A little weaker than I would have expected from ISM and factory orders. Its not that mfg is that large a factor anymore by itself but it is a 'driver' of activity in other sectors... mfg activity create or support lots of other jobs in services, transportation & such.
I would say the ISM survey might be bogus but not BOTH orders & survey. Maybe there is a lag.
The report looks even worse if one focuses on the household survey. In my Angrybear tradition, I graph the employment to population ratio - which alas dropped to 63.0%.
"The expected job losses in residential construction employment still haven't happened"
Again, I think the share of ilegals in the economy is not manifested in the stats.
The sales down figures from target, wall mart, sears, circuit city - all point to a sharper decline.
And K-mart. How could I forget K-Mart.
A bit softer than the expectations, but not significantly so. Like February, ADP nailed this number.
CR,
Does your forecast for hundreds of thousands of residential constr job losses over the next year or include the undocumented workers?
I am still not buying this talk of hundreds of thousands of job losses in residential construction. The jobs might appear to go but in reality many of the workers do find alternate work either in a more casual capacity by word of mouth or by doing something different where currently it is hard to find workers. The overall effect is therefore a softening rather than an over the cliff type of response.
Manufacturing employment declined, and was actually stronger than it should have been due to a strike that ended (accounting for 6,500 jobs). Finance and insurance lost 14,000.
Retail was down 26,000 net.
Did you see the results of the household survey? According to Marketwatch: "The separate survey of households showed employment fell sharply by 468,000 jobs in April. This is the biggest drop since November 2002. Unemployment rose by 77,000 to 6.80 million. The civilian labor force declined by 392,000." That is a sizeable drop, and a pretty big divergence from the establishment survey.
Unemployment rate among teenagers and Latino workers increase.
Homeowners Wage a Rebellion as Property-Tax Assessments Rise
homeowners-wage-a-rebellion-as-property-tax-assessments-rise: Personal Finance News from Yahoo! Finance
CR,
I remember at the outset of the recovery the household survey consistently "beat" the establishment survey in job creation. At the time, this was seen as evidence that the recovery was driven by a burst of entrepeneurial activity.
Could it be that, in reality, the household survey was better at counting residential construction workers? And that now we are seeing the reverse?
February & March were revised down 26,000.
And so will the April number be revised downward.
The household survey stat will certainly be ignored by the bulls, but even this payroll one is heavily revised, no?
The professional services contribute the most, yet the weekly wage actually moves down. Sounds like less hours to me.
Last nagging thing: the wait for construction jobs to disappear might just take longer or we face the music that the work performed by these "material movers" (as the BLS describes it) was not counted when it was performed and therefore not counted when it shifted to agricultural labor where it is seasonal, temporary and generally not "payrollable". There is a certain asymmetry about this undocumented insourcing and under/misdocumented outsourcing that keeps productivity numbers high.
Jed - Doesn't the household survey have much larger uncertainty? Order of magnitude with the actual 'signal'? I know the bulls usually trumpet household over establishment (at least the last few years as household was considerably stronger)... and bears usually complain about the error bars... isn't that the same problem now (but in reverse)?
I'd love to hear/read more detail on this... and not just gov't boiler plate... someone who could EXPLAIN the gov't boiler plate.
Is military personnel included in the government employment category. If so is the war contributing to the strength in the government employment numbers?
The business birth/death plug in the April payroll number was +317k, which is the largest ever for this statistical invention. I would think with a slowing economy this plug will prove quite optimistic (the April plug has almost always been revised down anyway) and the headline number will be revised down close to zero.
I've been more in the long period of very weak growth camp, but, in my mind, business confidence picking up (ism surveys, m&a activity) with the consumer coming under increasing pressure makes a hard landing more likely. This is because business is ramping up at precisely the wrong time, and recessions are generally caused by businesses scaling back inventories and capital investment (consumer spending rarely goes negative for any length of time).
calmo - you are 100% correct about productivity being overestimated due to poor man-hour accounting. I've talked with mfg wonks and they would agree strongly with your comment.
This is because business is ramping up at precisely the wrong time, and recessions are generally caused by businesses scaling back inventories and capital investment (consumer spending rarely goes negative for any length of time).
Turbo | 05.04.07 - 10:23 am | #
Could be but understand that companies don't 'ramp up' inventories very long anymore... couple very limited inventory space with tightly coupled supply chain JIT systems & there isn't anywhere to put the stuff. The warehouse is also the retail shelf.
If they have initially missed the signal - they won't miss it for long... the stuff will pile up in the isles if nothing else.
Watch consumption - that is the last link in the supply chain that stretches from the raw ore in the mine to the crap in your garage. Consumption is the last valve in the pipeline.
Couple wins $67,000 in lawsuit against predatory lender
SpringfieldNewsSun: Dayton, Ohio, news and information
The lender didn't have a license, misapplied payments, claimed she didn't make them and failed to pay her taxes out of escrow, leading to a mistaken foreclosure filing.
The expected job losses in residential construction haven't happened because the idiot builders are still building despite a gargantuan inventory of unsold houses.
Like so much of what is happening at the moment, this will delay the crash somewhat. Right now the rich, the powerful and the connected are frantically pulling every lever of power and propaganda in the hope of getting one or two more quarters of big bonuses.
But the crash is coming. And the longer it is delayed and denied the worse it will be when it inevitably descends upon us.
Dryfly - I agree. That's why the business cycle has become, well, less cyclical and that makes it hard for the gdp math to go negative (even for a bear like me) unless the consumer really craters. I guess the point I was trying (poorly) to make it that the current uptick in business confidence is unlikely to hold up for too long against a slowing consumer, and unlike the ravers trading stocks right now, I'd put much more weight on consumer measures, rather than business measures, as you suggest.
That being said, even though it is likely to be short-lived, a ramp-up in business activity now is just another overhang weighing down the economy by this summer. If we can go a week without another large buyout being announced, and the stock market stops going vertical, then the financial system will have to start dealing with its considerable problems, which will start the ball rolling in a way that makes me very nervous right now.
MSFT and Yahoo:
Are we to conclude that Yahoo was "undevalued" ?
Or maybe it is MSFT that is over valued and looking for a way to use that over valuation ?
In 99 there were many tech M&As....
"This is probably just the beginning of the loss of hundreds of thousands of residential construction jobs over the next year or so."
Never mind that you've done little more than meticulously back up this projection with comprehensive statistical analysis -- Your conclusion remains utterly irresponsible!
(Just trying to provide a public service here ...)
Dear Host,
I recall that the red line in earlier iterations of the top graph were more or less dead center in the range of expected outcomes. This month, the red line has come back into the expected range, after an excursion into happy territory. Am I remembering rightly? If so, what has changed about the graph?
Dryfly,
I wouldn't put too much emphasis on one month's headline factory orders data for guidance on hiring. I don't think the correlation is all that high. Moreover, factory orders are down 1.3% y/y, and orders ex-transport have only risen in 3 of the past 9 months. The factory hiring diffusion index is at its lowest level since late 2003. If I were doing the hiring, I wouldn't look for more workers under those conditions.
Steve,
Quite right that ADP has improved in its new version. I would note, though, that ADP anticipated fewer factory job losses in April, and that isn't what happened. Still, it's good to see improvement.
I wonder what employment statistics are for the real estate industry in and of itself. How many agents can keep hanging on when they have to share a smaller and smaller pool of buyers and sellers? At potentially lower prices?
If anyone has stats on that, it'd be interesting.
dryfly said: "I'd love to hear/read more detail on this... and not just gov't boiler plate... someone who could EXPLAIN the gov't boiler plate...
The boilerplate isn't really that intimidating.
http://www.bls.gov/lau/lauhvse.htm#hvse
I use the establishment survey because the data is "harder", less volatile and has a dramatically larger sample-size than the household data. I also use it consistently and exclusively over time, regardless of whether it supports my position or not.
Sebastia
The jobs might appear to go but in reality many of the workers do find alternate work either in a more casual capacity by word of mouth or by doing something different where currently it is hard to find workers.
Worried, that's the textbook definition of under-employment. It has the same depressive effect on economic activity.
Sebastian, you have your work cut out for you today.
But don't despair, you have plenty of like minded folk in the captive financial media. The bad job numbers simply mean the Fed has less incentive to raise interest rates. Of course, that also poses a problem for people who care about a weakening dollar. But then again, a weak dollar is a good thing, too, because that means U.S. internationals earn more $ on foreign exchange.
In the meantime, however, things aren't looking so rosy for those of us on Main Street as we watch our jobs go away and our purchasing power evaporate.
Round and round we go.
Kash has a great graph of employment growth by sectors:
The Street Light: April Employment Report
"The expected job losses in residential construction employment still haven't happened"
Is that before or after the +317,000 jobs made up by the B/D model?
Birth Death Adjustment: A whopping 317k B/D adjustment -- that is the single largest "adjustment" on record for any single given month.
Quite right that ADP has improved in its new version. I would note, though, that ADP anticipated fewer factory job losses in April, and that isn't what happened. Still, it's good to see improvement.
Yeah, some of the internals didn't quite match and they predicted large downward revisions for March, so their March number is even more off. Still, it looks like the recent changes improved things a bit.
Sebastien - Intimidating BLS it isn't. Incomplete it is.
I own/run a small biz - mostly family members dropping in & out. The wife & I fill out the establishment survey every quarter so I know what they are asking - it can't tell the whole story 'cause they don't ask the right questions.
In our case revenue is 1099 from clients to my company & then either salary or 1099 to us individually. We are not Sub S but run like a Sub S so it is a straight pass through & avoids double taxation. We pay the taxes individually like a partnership.
Most of what we 'do' doesn't show up immediately in establishment - I do but that's about it, rest of the 'work' folks do for me I 'contract'.
And believe me they are free agents & part-time... when you work with family cats start looking like reliable alternatives.
Many of the people I know are in the same boat. Even some jobs that used to be 'salaried' are now contract... For example I am having discussions with a couple companies now where I'll be their 'New Business Dev Director' - basically a hired gun hunter. Won't ever show up in their establishment anywhere... it will look like a 'job loss' on paper but isn't.
My daughter might work with me on this (she's a double engineer - marketing major)... she won't show up anywhere period either UNLESS I decide to 'hire her'... say to try & get her on a better insurance package under my company. If that happens then she'll show up on my establishment - probably not for many months afterward.
Eventually - most all eventually get captured in establishment somewhere but it is a real slow process. And they are just as likely to not be working for by then!
Like I said I know MANY people doing this - companies benefit by getting us off their benefits & pension roles & out of their offices... We benefit by not having bosses ride our ass about how & when we do what we do.
Of course getting paid is sometimes an issue - but that's 'off topic' here.
This trend is especially common in PEs & start-ups where the whole damned office might be 'equity holding' 1099 indy contractors. Lord knows there are a lot of PEs out there now.
But the BLS numbers don't accurately reflect this growing pool of labor - household tries to capture it but I don't think it does a great job - and this pool is VERY volatile, believe me, so it will never be an easy thing to do anyway.
BTW - I could just as easily see the trend to 1099 reverse IF pension & health care were no longer company responsibility... Execs hate working with indies like me, I'm also a bit like a cat that way... If I'm not making money then I'm blogging, or fishing, or watching grass grow... I don't do meaningless meetings nor cubicle time. Haven't for 20 years now.
Windshield time getting to and from clients & customers... or smokestack chasing... still doing lotsa that. Cubicles, no.
Today NFP report suggests that the next recession can start 6 weeks from now:
95,000k new jobs in NFP report this Friday « The Theroxylandr in Flame
Today NFP report suggests that the next recession can start 6 weeks from now:
What time?
Oh, Dryfly, I couldn't agree more. This is one of the reasons why I still feel that this economy has a lot of longterm strength in it. Your account also fits a lot of banking, info, etc.
I think this is where the "business and professional services" comes in. It drives me nuts that I see no direct way to measure it, but this very real, very efficient, supports productivity, and, I am guessing, will cushion some of the negative impact from our credit excesses of recent years.
The expected job losses in residential construction employment still haven't happened"
A HB site in 1989 had lot more workers then the same site in 2005. I have wondered if the number of construction workers hired for the HB sector has been significantly overstated during this build up. BLS and others using outdated labor rates and not taking into account new technology and prefab aspects of the HB trade today.
" I have a customer who's selling her condo and renting instead of buying because it's cheaper."
Business: A chilly June forecast for home sales
Steve, monday morning, as usual - I can't say more accurately.
Seriously, a lot of stuff is pointing that recession can start as early as July.
But NFP was the last shoe to drop. Before today, payroll data suggested that recession is impossible in the next 2-3 months. But today datapoint suggests that July is possible.
Let me stress that I do not conclude recession from NFP. I conclude from everything else, but NFP is sort of show-stopper, which sais, for example, that recession is almost impossible, say, right now in May.
dryfly, thanks for the response. I completely agree with you that economic data is going to be incomplete and have flaws (as does virtually all economic data). The subsequent revisions make things even harder.
Yet it's the best we have, and I've found it to work well...as long as I'm aware of its limitations, what it is and isn't telling me. It beats guessing and it sure as Hell beats massaging the data to get a pre-determined answer.
"...Like I said I know MANY people doing this - companies benefit by getting us off their benefits & pension roles & out of their offices... We benefit by not having bosses ride our ass about how & when we do what we do..."
So, you're saying MANY people you know had their jobs outsourced, but those jobs stayed right here in the U.S.?
Sebastia
I have never seen this before.
Both HOV and BZH had bad news released and both stocks went down.
I thought it is the other way around. Bad news = hitting bottom and the stock should go up.
Fake Jobs
The government numbers were fudged upward by the largest amount yet in the birth/death fake job adjustment. They 'added' 317000 jobs that actually don't exist except in the BLS statistics department.
Don't believe me? Look at the link and study the methodology. It is the ultimate government fudge factor. Skew any bad data into good data.
Don't believe the government's numbers, they are lies.
The real fact is that 88000(reported)-317000(fictitious)=229000 jobs were lost last month. Yes, you read that right, the real numbers were a LOSS of 229000 jobs in April.
BLS LIES!!!! Very shameful lying government data.
That is a VERY BAD jobs report, no way around it.
I think we are actually already in recession.
To continue somewhat with dryfly's description of white collar work that is not captured in the current employment surveys, I know several people including me in the blue-collar end that punch a clock for part of our income and then do side jobs to make ends meet.
Now the plural of anecdote is not data for our data-driven host but it does make me wonder if the current economy has changed enough to make revisions to the employment survey data collection methods necessary to accurately describe the current situation.
It seems to me our economy, despite the rabid growth in documentation for the simplest task, is much more informal than in past decades. And that informality may ignore the formal data collection process.
Jed wrote:
"The separate survey of households showed employment fell sharply by 468,000 jobs in April. This is the biggest drop since November 2002. Unemployment rose by 77,000 to 6.80 million. The civilian labor force declined by 392,000." That is a sizeable drop, and a pretty big divergence from the establishment survey.
Exactly, see the birth/death adjustment lies to understand the divergence, and why our present government needs to be eliminated.
dotcommunist, I think Hugo Chavez Fidel Castro and Jimmy Carter agree with you.
Steve, good point on the ADP.
ADP: 64K Private Sector, BLS: 63K.
pgl, nice graph, and I hope policymakers are paying attention to the participation rate.
David Pearson, I prefer the establishment survey for total employment, but the household survey is very good for unemployment (and participation rate as pgl noted). The slight increase in the unemployment rate is insignificant - but the participation rate is important.
Best to all.
barely, don't forget Rafael Correa, Michelle Bachelet, Daniel Ortega, Evo Morales, Hu Jintao, Segolene Royal, Ándres Manuel López Obrador, and a hell of a lot more people in the USA than you would obviously like to believe.
Keep dreamin' for the past, barely.
...though I don't really see what your comment has to do with the government manipulation of job statistics.
Ordinarily dotty, I'd tell you that the birth/death model(s) have worked well over the past few years as evidenced by smaller subsequent revisions, but today I'm feeling less striped and more dotty...and I think their methods are not tested for significant trend changes and you could be spot on...but maybe only this time.
Workers' paycheck also grew more slowly.
Over the past 12 months, wages grew by 3.7 percent
Jobless Rate Rises As Hiring Slows
I'm confused, Steve and the other permabulls keep repeating the claim about that strong income growth, but somehow, I just keep seeing the opposite.
Calculated Risk, I want to suggest this answer to the conundrum.
The data are cloudy. But based on arguments I have made on my blog, I think that we are seeing real job growth in health care, probably low wage. As for manufacturing, there's too much internal inconsistency between the establishment survey and ISM to believe that factory performance is actually improving. Therefore, my guess is that factory orders did not significantly rise and that next month this will be evident.
The reason this is relevant is that health care spending on care for the elderly is another major economic problem the US faces. If employment for hospice care rises, it comes out either Medicare or savings: more debt on the consumer or on the government and therefore lower future growth. Whereas rises in civilian factory orders would probably imply an improvement in the trade situation.
There, I think, is the bull v. bear case in a nutshell. And so it all depends on where jobs were actually created, on the factory floor or in home health work.
Looking at the seasonally adjusted employment-population ratio:
previous peak:
4/00 64.7
previous trough:
9/03 62.0
current peak:
12/06 63.4
current:
4/06 63.0
Job quality is also something that isn't tracked with job creation figures: full time or part time? Contract, temp, or perm? Low-wage or high wage? Benefits, or no benefits?
Roll in all that and you might have the bare beginning of a meaningful metric as to the health of the job market. The ones we have are so much marsh gas.
dotcommunist said: "I'm confused, Steve and the other permabulls keep repeating the claim about that strong income growth, but somehow, I just keep seeing the opposite..."
Tell me about it.
Sebastia
Indeed, mynianville also talks about NFP report to be udjusted upward by 317k by birth/death model.
Five Things You Need to Know: Economic Deceleration Contained to Overall Economy; GM Losses Contained to Auto Sales and Subprime Lending; No Worries!; What Exactly Is a Moratorium?; Hey There, Little -Minyanville
Which is the biggest adjustment as far as the table show.
It seems that the actual NFP report is, probably, NEGATIVE. Welcome to recession within 2 months!
Didn't someone (CR?) show that the birth/death adjustments don't work well during economic transitions?
theroxylandr,
While I agree that the birth-death model is a suspicious black box, the 317k is NOT seasonally adjusted and the not seasonally adjusted employment increased by 833k. So without B/D model you have 516k new jobs in April. I have no idea how to apply seasonal adjustment (I believe the seasonal adjustment is applied separately to each sector and the B/D model is also applied separately) to this figure but you cannot simply subtract 317k from 88k.
Can somebody point me to the birth/death number in the release. I can't find it
TJ, from the bls birth/death model page,http://www.bls.gov/web/cesbd.htm
"The most significant potential drawback to this or any model-based approach is that time series modeling assumes a predictable continuation of historical patterns and relationships and therefore is likely to have some difficulty producing reliable estimates at economic turning points or during periods when there are sudden changes in trend."
Basically, there's little way to figure out what these figures mean for the overall economy.
Lower employment figures MIGHT be a positive - reduces inflationary pressures. They also could be a negative, in that the economy is slowing.
But I usually ignore employment figures when trying to figure out where the economy is going for two reasons.
1- It's mostly a backward looking statistic.
2- If you go to the museum on Ellis
Island there's a great graph showing immigration over the past 200 years or so. There's a very marked pattern of reduced immigration and even emigration during "temporary downward periods of economic activity" (AKA recessions, depressions and panics). My grandmother, for example, was born in the US but moved back to Poland with her family during the recession in the early 1920s and then back to the US in the late 20s boom.
There's been a historically large amount of immigration over the past decade, or at least there has been in my neighbourhood - Manhattan is close to its highest ever levels of residents who were born overseas. Ive always suspected variance in immigration/emigration is important and isn't captured in the stats. I was very glad to see a similar point being made in a NY Times article last week.
If the B-D model is an extraordinary sham, it will be exposed in a year or so. Missing jobs show up in plenty of other stats.
I don't like the idea that the scorekeepers would suddenly yield to pressure and produce worthless numbers.
Pressure is always there, as the cycle is always cycling, roughly once an election interval. The scorekeepers jobs are relatively secure.
If the argument is that all of the numbers are a sham, then whats the argument? Pick up your ball and go home.
More to point for this blog, the birth-death model added in 49,000 (not seasonally adjusted) construction workers in April. That's more than a fourth of the 188k reported in the release. It seems that in prior years, April was a pretty good time of year to start a residential contracting business. But the birth-death model has no way of knowing that those business births didn't happen this year, and in fact, there were probably quite a few mom-pop operations that decided not to build any homes this year.
The birth-death model is not a sham, nor are they "fictitious" jobs. But they are estimated jobs that will either be proven to have existed or be proven not to have existed once the hard numbers come in from the data like tax returns.
Most likely, when it's all said and done, we'll see construction jobs fell by at least 50,000 in April.
The bottom line: the report was very weak EVEN WITHOUT a huge decline in construction jobs that will likely show up in the revisions.
Dryfly, when I was contracting as an engineer years ago I noticed a pretty tight coupling between the direction of the prevailing economic winds and the availability and duration of contract jobs. They were usually the first cut when things started edging south. Have you seen any such slowdown out there?
Nice review fred. I also like The Nabob's view of some of those details:
The Nattering Naybob Chronicles
IBM to lay off 100,000 in the US?
I, Cringely . The Pulpit . Lean and Mean | PBS
A bit speculative, but Cringely claims to have some inside info...
The idea that BLS is somehow distorting the numbers is rather wild to me. I don't think the American taxpayer wants to pay for BLS surveying the entire population of households each month.
They do the best they can, and then they correct with additional data. The numbers are not going to be perfect.
I'll check in with one of my cousins on the IBM thing.
check out
IBM Employee News and Links | December 5, 2009
wow!
Dryfly, when I was contracting as an engineer years ago I noticed a pretty tight coupling between the direction of the prevailing economic winds and the availability and duration of contract jobs. They were usually the first cut when things started edging south. Have you seen any such slowdown out there?
Andrew - I am the most counter-cyclic contract worker in America... I do better & am in more demand in busts than in booms.
My job is to find 'new business' for mfg companies and when they are full of work & have backlogs growing the last thing they want is 'new business' Then they tell me to go away. Not all that politely either.
Over the last year though I've been contacted more regularly but surprisingly not as much by US firms. Many of them have been consolidated into large PE takeovers and they don't need indy hired guns.
When the PEs buy a new firm it has a Sales & BD staff already in place so cherry picks the staff of the new firm for who they want then toss the rest out - they sure don't need more.
But offshore suppliers to US firms are really stepping up their presence here. They don't just want to work for the US owned FDI operations in Asia - they want to do what the Japanese did - go fully global, here as well as there.
People like me already know the lay of the land so they contact us - I listen to them, believe me.
Three of the four best new prospects I have are from outside the US... two from India, one from Canada... only one was a US firm (a brand new start-up).
I am reluctant to work with offshore accounts - too hard to make them pay you if they decide they aren't really interested... And who wants to sue them in Bangelore?
But that might be what I have to do if the US firms continue to consolidate like they have over the last couple years. If you can't join'em I guess you fight'em.
Tell me about it.
Sebastian
Don't worry, we all know whose side your on.
Reality