As a follow up to our NFP extravaganza yesterday, here are a few additional items you may find informative:

Yesterday, we noted the big jump in Unemployment (a historic outlier) and wondered aloud whether it might be attributed to, in large part, a seasonal aberration of the younger set, the teens and 20-25 year olds.

As it turns out, not so much: Philippa Dunne and Doug Henwood of the Liscio Report have crunched the numbers, and they suggest the impact of Teens is de minimis. In an email I received late last night from Philippa, she writes:

"Teen unemployment rose 3.3 points, which was probably exaggerated by some calendar issues. But teens are less than 5% of the workforce, so they contributed just 0.2 point of the total rise. Adult unemployment rose from 4.5% to 4.8% – and it was a clean move, with no rounding funniness. The adult participation rate rose 0.1 point, and the EPR [Employment Population Ratio] fell by 0.2 point. Also, the composition of the unemployment rise shows that it wasn’t just the kids: permanent job losers and re-entrants accounted for 0.2 points of the rise, and new entrants just 0.1 point.

If you want to spin a story out of this, it could be that people are re-entering the labor force because they’re having a hard time making ends meet."

-Liscio Report

Thank you, Philippa — I stand corrected on the teen portion of NFP.

Downward Revisions were also noteworthy. March was revised down by 7,000, and April by 8,000. This was the the fourth consecutive month of downward first revisions, as well as the fourth consecutive month of second  downward revisions. It is further evidence of the negative overall employment trend we discussed  in our NFP Preview — weak private
sector job creation; poor real wage gains; flat to negative hours worked.

Last, the Birth Death Adjustments: As we noted yesterday,  42,000 new jobs in construction and 9,000 new financial jobs both seem rather improbable.

We continue to see evidence that in the latter half of a cycle that the B/D overstates job creation significantly. The chart below shows the distortion the Birth Death portion of NFP has created: The B/D jobs have remained
steady, while the actual measured portion of NFP net job creation has been faltering since early
2006:

>

Blsbd

chart via Jake

>

If the BLS is going to continue to use the Birth Death adjustment to capture new job creation the establishment survey misses in the beginning of cycles (due to new firm creation), then they need to figure out how to make adjustments for the declining portion in the latter portion of the cycle.

Until the business cycle is repealed, the Birth/Death adjustments (similar to how seasonal adjustments are neutral over a year), should be more balanced over the full course of an economic cycle.

Category: Data Analysis, Economy, Employment

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

25 Responses to “More on May 2008 Employment”

  1. Scott Frew says:

    In an email I received late last night from Philippa, he writes:

    Uh, Barry, last I knew, Ms Dunne was a woman, and deserving of the pronoun “she”.

    Scott

  2. Doh! I’ll fix above.

  3. Carl says:

    The ‘disappoint’ of the NFP number had ramifications elsewhere.

    The following shows what the Sell Programs looked like in the pre-market when the number was announced at 8:30 and how they dominated throughout the day.

    http://www.transactionlevelanalysis.com/2008/06/the-crash-of-87-what-if-you-ha.html

    You may find the graphic interesting as chart porn.

    Dig your blog…alot.

  4. Jim Haygood says:

    “We continue to see evidence that in the latter half of a cycle that the B/D overstates job creation significantly.”

    Ditto. B/D has the characteristics of a linear extrapolation. Works fine during a steady-state trend. Breaks down at an inflection point, such as a transition from growth to contraction.

    Does anybody ever test this stuff? Linear extrapolations can be applied to market forecasting. But they work no better than any other trend-following technique. Which is to say, the “edge” is marginal at best.

    But given the uses to which employment data is put, I don’t understand why one would want to pollute the data with a linear extrapolation which (in effect) introduces a lag at turning points. This is supposed to be coincident data, not a forecast.

    And I’ll bet you that even if we got a three-year long, capital-D Depression, a la 1929-1932, the Birth/Death model will NEVER be used to subtract jobs in an ongoing contraction. What do you think these are — UNBIASED STATISTICS?? LOL!

    If they ever did, of course, it would be the greatest contrarian buy signal of all time — even stronger than a “Death of Equities” magazine cover.

    Conclusion: situation normal; all fubared up. Nobody here but us value subtractors!

  5. grumpyoldvet says:

    And yet there on Kudlow’s show was Westbury, Bowyer and Luskin spouting the White House propaganda line….http://rawstory.com/news/2008/CNN_reporter_Bush_excuse_for_unemployment_0606.html
    When will CNBS get Kudlow off the air and send him packing….

  6. John Borchers says:

    There’s only one way out of this negative spiral as I see it. Americans need to, whenever they can, purchase American made goods.

    In 2006 I bought a Nissan. 2 years and 21k miles later it has body and clutch noises. This certainly isn’t the same quality it used to be. After I bought it I found out the cars are assembled in Mexico. Doh!

    My point is I probably would have been better off buying an American made car since large amounts of the business are outsourced anyway and I would have helped keep some US people in work.

    So from now our family will try to support US companies and deny imports as much as possible.

    I’ve been buying F stock this week. Something I’ve never owned before in my lifetime.

  7. dukeb says:

    “…it could be that people are re-entering the labor force because they’re having a hard time making ends meet.”

    Now *that’s* one of the most interesting thoughts I’ve read in a while. Let’s see how it runs.

  8. grumpyoldvet says:

    John, it’s virtually impossible to “buy American” any longer. We have outsourced our manufacturing base a long time ago. We are for the most part an assembler. Even many services today are being outsourced, think accounting service, tech and info support, medical diagnosis, financial analysis, etc. Yes there are some businesses that are strictly American, but they are few and far between. The economists will tell us we need to become an “educated” nation to compete but even education is becoming difficult for many middle class families and forget those with less, they are screwed.
    I often told my father, a WWII vet, that his “greatest generation” took all the goodies and left us with the bag. Same with my generation, that VN era group, we took ours and left our kids and grandkids with the bill

  9. John Borchers says:

    I realize it’s not easy Grumpy. But when push comes to shove if we have a choice between buying a US companies’ outsourced product like Nike for example against just some rinky dink shoes made in China it would be better to buy the US Nike as at least we will be supporting some US jobs.

    I think there will be quite a switch over the next generation. I see manufacturing coming back to the US from abroad.

  10. VennData says:

    Adjustments can’t catch secular trend changes. Compile the raw data and let everyone who wants to have their own – branded – numbers with what ever adjustments they want. Brian Wesbury can have his, bloggers can have them, the U.S Chamber of Congress can have theirs. etc… then, over time the market will sort it out.

  11. blowhard says:

    it’s very funny to me that the bobbleheads have tried to downplay the unemployment stuff by basically saying “ahhh, don’t worry, it’s just the teenagers that can’t get work.”

    i would imagine there’s a sound argument to be made that these “teenagers” are far more important to the discretionary-sensitive consumer companies than anything else. Most of these “teens” probably aren’t that good at starting out their IRAs and just spend the money as quickly as they make it, now there money is more likely going to gas and food, or helping out the family instead of buying discretionary things.

    As cool as Apple is, it’s an image/trend based consumer electronics company, not a corporate tech company. Those ugly furry UGG boots the hot little college girls all seem to love, where’s that money going to come from if the “teens” can’t get their $7 an hour.

  12. Jim D says:

    IIRC, the best odds of buying American is to buy Toyota.

  13. grumpyoldvet says:

    Nike sneakers…..ah there’s a product…what do they cost $125/$150 per pair…ever wonder what the Chinese, Cambodian, wherever laborer gets for making those “American” sneakers….

  14. Mark says:

    re: B/D adjustment.

    I nearly fell out of my seat after the ADP report on Wed. when the talking head from Macroeconomic Advisors said that the ADP report has been diverging to the high side recently from the BLS numbers because the BLS B/D adjustment has been overly pessimistic!

  15. wally says:

    There is no justification for using the ‘birth-death adjustment’. If they believe their number incorrectly shows the actual total, they should just say so. Or, they could simply use a multilpier (which is what birth/death is anyway). In any case, you don’t make something more “accurate” by adding a guess to the hard numbers. Bigger, maybe, or smaller, but not more accurate.

  16. John Borchers says:

    Jim,

    If you are buying anything long in the stock market right now you want to buy what is already selling at depressed rates. These are the mall retailers (JCP, M), American car companies (F,GM) and the home retailers (HD, LOW). I’m avoiding banks, brokers, home builders as current values I don’t believe are depressed enough. Bank profits will be cooked for years along with home builders. The others will start to make recovery early next year.

    I want to buy what has been hurt by high inflation of the commodities.

    I want to buy after all hopes that the economy will come back in the back half of 08 get smashed to pieces but that means I need to start buying now, and that’s what I’m doing.

    My short positions are all sold, including LEH.

  17. tranchefoot says:

    Interesting post. I wonder how many of the newly unemployed are just already-fired workers who’ve busted through their severance. If anything, that trend is going to accelerate.

    BTW, Barry, did a post from last night disappear? The one about the brokerage analysts?

  18. tranchefoot says:

    I don’t know, JB, those stocks MIGHT be considered inexpensive, but they definitely aren’t cheap, and I still think it’s too early to buy. There are still a lot of people out there who don’t think we’re headed for a recession–notwithstanding yesterday’s wake-up call, there are still Kudlowians out there who probably won’t see a recession until we’re three months finished with it, and even then only if there’s a Democratic president at the time–and you don’t want to buy before people are depressed about their portfolios.

    And there’s still questions, I think, about whether or not GM and F can even SURVIVE a recession.

  19. Mich(^IXIC1881) says:

    I’ve been buying F stock this week. Something I’ve never owned before in my lifetime. Posted by: John Borchers | Jun 7, 2008 9:25:05 AM

    – $165B is a mighty amount of debt to be financed at favorable rates. I bought puts on F as a hedge a while back. Living in Detroit, I want to make sure if/when F seeks bankruptcy protection, I have at least some good news coming out of it, because it will devastate the local economy. I sold it on Friday though, but still want to get into another hedge in the near future. I wish I knew about USO a while back, i would have bought that as a hedge to my gas costs as well. too late now.

    John, it’s virtually impossible to “buy American” any longer. We have outsourced our manufacturing base a long time ago. We are for the most part an assembler. Posted by: grumpyoldvet | Jun 7, 2008 9:58:44 AM

    – GM/F proudly presents vehicles that are fully assembled by the “made in the USA” screwdrivers.

  20. ScottB says:

    I believe once again you have misunderstood the birth/death adjustment. As comments in the past have pointed out, the B/D adjustment has two components: trend and seasonal. The trend part of b/d is picked up from the sample of existing firms, and is not published. The b/d trend, during a downturn, is of course negative. The seasonal part of the trend (the number that is published) for construction is ALWAYS positive in the spring, regardless of the business cycle. The two added together estimate the net number of births and deaths.

    ~~~

    BR: I am familiar with their methodology. I do not agree with it.

    Regardless of the method to their madness, I simply do not believe 42,000 new jobs in construction and 9,000 new financial jobs were created in May.

  21. cm says:

    Barry

    The prior to last paragraph seems cut off — ending in ‘Like’.

    ~~~
    [BR: Fixed]

  22. cm says:

    blowhard: “where’s that money going to come from if the “teens” can’t get their $7 an hour.”

    Daddy’s/mommy’s purse, as long as that lasts. Parents will generally “enable” their kids’ peer comparisons, subject to financial constraints. Has it ever been otherwise?

  23. Estragon says:

    ScottB,

    I’m not entirely clear on the seasonal element to the B/D adjustment. From the BLS faq :

    “Q: Are birth/death factors seasonally adjusted?

    A: No, they are calculated using population data that is not seasonally adjusted and the factors are applied to the sample-based not seasonally adjusted estimates. Months with generally strong seasonal increases such as April, May and June generally have a relatively large positive factor. Conversely, months with overall strong seasonal decreases, such as January, generally have a relatively large negative factor.”

    In any case, the BLS also says:

    “the first component excludes employment losses from business deaths from sample-based estimation in order to offset the missing employment gains from business births. This is incorporated into the sample-based estimate procedure by simply not reflecting sample units going out of business, but imputing to them the same trend as the other firms in the sample. This step accounts for most of the net birth/death employment.

    (emphasis mine)

    The assumption underlying this method is that workers essentially move from previously sampled dead businesses to not yet sampled new businesses in numbers equivalent to the state of existing businesses.

    Most of the time, this apparently fits the facts. At major turning points though, the assumption may not hold.

    For example, it’s likely very few new construction businesses are being started now, and a lot of smaller weaker contractors have folded. Some of those workers may have been hired by larger contractors who are still building multifamily and commercial projects. Continuing firms will appear to have a positive trend, but the new firms that in normal times would be replacing the dead ones aren’t really there, but are still modelled in based on the growth in survivor firms.

    Construction is probably more prone to this sort of problem because trades tend to go out on their own in boom times, but can easily move to larger firms in busts. In other words, the structure of the industry adapts quickly to cyclical change. The structure of (eg.) retail will change much less radically in a downturn, in that continuing firms are likely to show job losses more proportionate to the change from births/deaths. In other words, the structure of dead/continuing/new firm employment will be relatively stable and better reflected in the B/D model.

  24. Puzzling Dragon says:

    cm and blowhard, you both make good points, and in a belt-tightening economy, both sources of discretionary funding (summer jobs and economic outpatient care) will shrink. Perhaps they will not shrink to zero, but the narrowing of those funding pipelines will still affect companies that rely upon discretionary consumer spending… probably in a fairly substantial manner.

    -pd

  25. blowhard says:

    cm-

    Maybe that’s what happened in this job report…