Its tine for everyone’s favorite monthly data point, Non-Farm Payrolls.

For those of you who are relatively new to the site, here is a brief description of how our analysis has evolved over time.

Weak Jobs Recovery: Following the 2001 Recession, the economic recovery, from a jobs perspective, was rather weak. Indeed, from 2002-07, we had the weakest employment recovery of any post-recession period since World War 2. This data point — widely ignored on Wall Street and MSM — was a warning sign that the recovery was abnormal. It is what sent us looking for what was driving the economy — and the answer was borrowed money.

Survey Data: There are two employment surveys — Establishment and Household. Establishment works of employment tax data; Household is literally a Q&A survey. They sometimes vary dramatically, but when you control so they measure the same thing, they come pretty close to each other (most of the time).

Birth Death Adjustment: A major modification to the NFP measure is the Birth Death adjustment. Changes to the BD were proposed in 2001, and implemented a few years later. This attempts to capture early improvements in employment at the start of a recovery was the goal. The trade off is it wildly overstates strength at the end of a cycle. For example, in 2007, approximately 75% of reported new jobs were due to this adjusatment. In 2008, the BD adjustment inexplicably showed lots of job creation in construction and finance.

Measuring Unemployment: The main measure of Unemployment is the widely reported U3 Unemployment Rate — but my analysis of U3 has been that it significantly understates unemployment. Fortunately, a more complete measure of labor under-utilization is available –  the U6 measure. They seem to run parallel, but U6 captures a lot more of the unemployed and under-employed workers than U3 does.

Leading vs Lagging Indicators: Lastly, economists will tell you that Employment is a lagging indicator, meaning that it lags the economic cycle, getting worse even after the economy begins to improve.  And that is mostly true. However, since we are investors by trade, we want o identifty aspects of Employment data that have the qualities of a leading indicator — i.e., they improve before the economy does.  There are at least 2 worth paying attention to: Temporary Help, and Hours Worked. Both aspects improve or worsen prior to the a recovery or recession occurring.

Note that there was a lot of pushback against these ideas when they were first  discussed here; They have now more or less been recognized if not endorsed by many astute observers . . .

~~~

Employment Situation report out a 8:30 today; Bloomberg consensus is 365,000 jobs lost; Barrons Consensus is 350k witha  range of -435,000  to -225,000.

Category: Data Analysis, Employment, Wages & Income

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.

50 Responses to “Explaining Non Farm Payroll & Analysis”

  1. I’ll see if I can dig up relevant links to these prior discussions . . .

  2. Tony61 says:

    I probably should know this, having read this site for a number of years…. my understanding is that the birth/death adjustment has to do with “birth of new companies”, and there is some type of implied rate of new company formation factored in. Is this correct? Can someone explain this better or provide a reference. Thanks… and thanks Barry for all you do.

  3. dead hobo says:

    Is the BD adjustment removed over time as the revisions dribble in? Or is the error compounded, like interest over time?

  4. cvienne says:

    ” Indeed, from 2002-07, we had the weakest employment recovery of any post-recession period since World War 2″

    IMO – The “recovery” was not a recovery at all (except in equity prices)…

    The whole period was a big drunken slobfest where Americans pulled counterfeit money out their artificially inflated homes…The only reason we’re still standing now is that now we’re pulling counterfeit money out of the Treasury…

    We’ve been in a bear market since 2000…This bear ends in late 2011 (at the earliest)…

    NFP will reflect that…

  5. dead hobo says:

    cvienne Says:
    July 2nd, 2009 at 8:11 am

    ” Indeed, from 2002-07, we had the weakest employment recovery of any post-recession period since World War 2″

    IMO – The “recovery” was not a recovery at all (except in equity prices)…

    reply:
    ————–
    Well, I’m not giving the money back.

  6. call me ahab says:

    jc-

    in a natural resale market- the occasional distress sale would get ignored assuming there is an abundance of normal resale activity- however-

    your observation is correct- when short sales and foreclosures comprise a large percentage of resale activity- then- they are driving the market and prices and have to be accounted for in an appraisal-

    I see the 125% refinance as having ZERO impact- and will only attract a select few that may fit into its parameters- first one being- that the owner wants to keep the house- as has been noted before- even those folks who had mortgage modifications end up re-defaulting – because they don’t want the house- and if they could they would wave a magic wand and make it all disappear

  7. U.S. employers cut 467,000 jobs in June. The unemployment rate rose to 9.5% from 9.4%

  8. cvienne says:

    @DH

    “Well, I’m not giving the money back”

    Yeah baby…I guess we’re all starting to figure…There’s more where that came from! Money apparently does grow on trees…Who needs a job?

  9. cvienne says:

    @ahab

    “I see the 125% refinance as having ZERO impact”

    Oh it’ll have SOME effect…The effect it will have is that after you do the re-fi, it buys you about 6 more months to stay in the house without making payments before they kick you out from defaulting on the RE-FI…

  10. HCF says:

    -457,000….

    Larger than the WORST estimate from Barron’s….

    HCF

  11. AmenRa says:

    weekly hours decreased -.01 to 33.0
    U6 is 16.5%
    B/D added 220k

    Love those green shoots!

  12. call me ahab says:

    cvienne-

    agreed- that is what will happen- but those taking advantage will be a small sliver of the much bigger default pie-

    I feel poetic today

  13. manhattanguy says:

    Reality Barry Reality. I predict a double dip recession in the offing.

  14. Bruce in Tn says:

    Hours per week…..33.0….shoot!

  15. investorinpa says:

    Still trying to figure out how they expected 365K jobs and a 9.6% unemployment rate…but instead got 467K jobs and a 9.5% rate. There has to be a better statistic out there than this…for instance, how about totalling up overall wages and dividing it by number of hours worked, or number of workers? I just don’t get all the BS and how it slips by the media. I mean, they aren’t that dumb, are they?

    ~~~

    BR: 2 Different surveys — different methodologies — that answer each question separately.

  16. Transor Z says:

    Here’s the B/D explanation w/methodology from BLS.
    http://www.bls.gov/ore/pdf/st020090.pdf

  17. Mike in Nola says:

    Amazingly, CNBC had a pretty decent set of people on this morning.

    Current B/D adjustment numbers anyone?

  18. investorinpa:
    If you watched CNBC, you’d know that they were that dumb. Whether willfully or on purpose I don’t know.

  19. willid3 says:

    i always thought that we really never had a recovery from the last recession. we just had a credit fueled recovery. which blew up at the end of 2007. and is still collapsing

  20. manhattanguy says:

    I am a better trader since I shut off CNBC (cable for that matter). It’s pure noise.

  21. dead hobo says:

    Transor Z

    Thanks.

    Excerpt from report follows. I don’t think the bias is removed over time. However, they appear to recognize inaccuracies at economic turning points. On balance, it’s probably horse shoes.
    —————————————
    As a final note, the most significant potential drawback to any model-based approach is that time series modeling assumes a predictable continuation of historical patterns and relationships. Therefore, a model-based approach is likely to have some difficulty producing reliable estimates at economic turning points or during periods in which there are sudden changes in trend. With the net birth/death model, this difficulty is significantly reduced, as imputation allows for such trends or turning points to be partially captured in the estimates. These imputed amounts are based on the active, reporting establishments which reflect changes in the economy. Therefore, contributions from the imputation for business births will also vary during economic changes. Furthermore, the imputation error from the net birth/death model is forecasted from the errors of the last five years. The error amounts from the most recent year are weighted more heavily, thus allowing recent changes to be more effectively captured. Even so, accurate estimation of the business birth component of total nonfarm employment will continue to be the most difficult issue in CES employment estimation.

  22. Mike in Nola says:

    Was gonna watch Dylan this morning. The Reverend is on, talking about Michael Jackson. So, I guess no go on that.

  23. Bruce in Tn says:

    Franklin:

    This is why investing on hope is dicey…It is ok to be optimistic, but the idea is that there is 12 trillion less of wealth in America than there was in 2007…that is almost one full year’s GDP….

  24. Steve Barry says:

    Who cares about these phony numbers…look around…where I live, I see hundreds of thousands of sq ft of commercial space sitting vacant and I can’t imagine when it will possibly be filled. Some of it has already been vacant for years. I have neighbors who I’m finding out don’t have jobs and they are pretending they do. I have a friend who bought a million dollar home and is now near foreclosure. the B/D model makes these numbers useless.

  25. Mike in Nola says:

    Looks like B/D is 185k this month compared to May of 220k. A bit less effect than last month. So, if you subtract the diff, you get about a 100k increase from May.

  26. grumpyoldvet says:

    The reported number showed decliines in all categories EXCEPT education & health care. Now I understand healthcare to take care of us old farts but an increase in education. So what do I make of this..

    http://www.nytimes.com/2009/07/02/education/02school.html

  27. rootless_cosmopolitan says:

    Barry, your implicit suggestion that the B/D-adjustment significantly distorts the real nonfarm payroll data, compared to an approach that didn’t apply the B/D-adjustment, needs to be proven using actual data. You only make the claim again and again in your postings. The benchmark revisions done every March don’t support this claim, though.[*] For instance, nonfarm payroll data were revised down only 0.1% in March 2008. This doesn’t support the hypothesis that that B/D-adjustment led to a significant overestimation of job creation in 2007. We will see what the March 2009 benchmark revision shows for the 2008 data.

    [*] http://www.bls.gov/web/cesbmtab.htm#Tab1

    rc

    ~~~

    BR: I thought it was pretty clear and understandable above, but I will clarify.

    In the beginning of an economic recovery, the BD captures new job creation that the Establishment survey (which focuses on larger firms) misses. The trade off is the B/D sees job creation later in the cycle when its not really there. And by jobs, I mean gainful employment including salary. Getting fired and becoming an independent contractor (unpaid) isnt the same.

    I have discussed the data repeatedly here (try using the site’s a search function). But here is an example:

    BLS Overstated Job Creation by 14.38% in 2007

  28. jc says:

    Brown shoots, even Kudlow will have some problems spinning this number

  29. Mike in Nola says:

    Steve:

    My interest in them is only how the market will react. As Taleb said this morning on CNBC, so I turned the sound on, a monthly number is pretty meaningless. Interestingly, he was pushing the housing solution Hussman has been for a year or more: banks adjust mortages to current value but get an interest in any upside appreciation when/if it occurs. Of course, as long as the Feds allow 125% LTV refi’s what incentive is there for the banks to write down loans.

    I am also seeing a lot more empty storefronts here in Houston inside the loop which is a higher priced area. Local real estate blog posted a little expose’ on it here: http://swamplot.com/all-that-empty-retail-on-south-shepherd-a-drive-by-photo-tour/2009-06-25/ The commercial area described borders what had been a pretty prosperous area.

    Santelli was wild this morning with the nflation rants despite the opinions from several qualified guests.

  30. pjl says:

    Those looking for additional sources of analysis on the jobs data should check out http://www.acropoblog.com

  31. cvienne says:

    @grumpyoldvet

    ” but an increase in education. So what do I make of this”

    You make of it as to understand why Franklin is so optimistic…Because this country will never live up to its potential until we drain the Treasury of money that’s not even printed yet to create a nation full of schoolteachers preaching the evils of anything that’s not ultra liberal…

  32. rootless_cosmopolitan says:

    @Mike in Nola:

    But what makes you think that subtracting the change from the B/D-adjustment gives you a number for the nonfarm payroll change that is closer to reality than w/ B/D-adjustment?

    rc

  33. willid3 says:

    dumb question. since the vast majority of small business do not have any employees, doesn’t the use of a birth/death model cause the errors? such as over counting the number of jobs?
    and wouldn’t they need to adjust them on a yearly basis when they do a count of real jobs that exist?

  34. Mike in Nola says:

    rootless: only doing a comparative, which is about all you can do. I am sure the margins of error are big. The trend is about the only use of these.

  35. ella says:

    With week after week of initial unemployment claims of 625, 000 how is that the monthly BLS data is coming in substantially less. Me, I am not buying the BLS data.

  36. uno says:

    Right side of Russell 2000 shoulder…arriving.

  37. Andy T says:

    I guess all that babble I was hearing yday on Bloomberg, via the Bespoke group, that a bad number would lead to a stock market rally (’cause traders would believe it couldn’t get much worse) was pablum…

  38. manhattanguy says:

    Dow and S&P both broke 50 ema. I expected this to happen on July 6th. Not complaining a bit though -:) My $DUG and $COF shorts are doing great.

  39. AmenRa says:

    Doesn’t a reduction of weekly hours by .01 equate to a job loss of 350k?

    The household survey counts the number of people who are/aren’t working in each house. The establishment survey counts the number of jobs gained or lost. One person can have more than one job which is a small part of why the surveys can have conflicting results.

  40. Mike in Nola says:

    Andy T: The computers haven’t had their second cup of coffee yet :)

    BTW, anyone see this video http://zerohedge.blogspot.com/2009/07/themis-trading-principal-program.html

    Someone else publish the disclosure people have to approve to use the Goldman trading platform. It basically allows Goldman to monitor users’ activity and frontrun them.

  41. Steve Barry says:

    Again, given the slaughter in retail I see everyday, small businesses are hemorrhaging jobs…since the B/D is adding close to 200K a month, the headline number is a joke…work week, temp hours and U6 are much more useful, if you can trust them.

  42. AmenRa says:

    @Andy T

    weekly hours > unemployment rate > # jobs lost. As long as the weekly hours stay the same the reaction is muted. But when the hours decrease AND the rate increases then we have a problem ;)

  43. Steve Barry says:

    @Amen:

    Good point…a business can cut a worker or save the same money by not firing someone, just cutting everyone’s hours.

  44. rootless_cosmopolitan says:

    Barry, I understand your hypothesis, which you have repeated in your reply to my comment. I even think there is some plausibility in what you are saying. But saying what you think what is happening is not the same as proving it using actual data. For instance, it has been pretty clear to me that you think the B/D-adjustment led to a significant overestimation of the nonfarm payroll numbers for the year 2007. But if this is true why were the nonfarm payroll numbers revised down only slightly in March 2008, when the benchmark revision was done? Maybe your hypothesis will be supported when we know the data from the benchmark revision of March 2009. It’s possible that the overestimation is not significant at the peak of the business cycle, but during the contraction phase.

    rc

  45. Todd says:

    Today is setting up to be a good fade the gap day.

    Nothing has changed, the only positive news I see coming out is that car inventories will be falling for GM and Chrysler.

    The real numbers are the Steady rate of 600k initial claims. The unemployment initial numbers are always adjusted. It has always under reported in the direction of the momentum.

  46. jc says:

    Off topic, what impact do you think upping F&F mort capability to 25% underwater (vs measly 5%) will have?

    How important will appraisals be in areas where there are a lot of 50% off foreclosure sales, do appraisers diregard those low values – even when they account for 1/2 the closings in an area?

    -125% only applies to primary morts so I’m guessing an awful lot of homes with second morts & HELOCs will get knocked out.I assume these are non-recourse so uncle sugar will really be on the hook bigtime

  47. Bruce in Tn says:

    I am a big fan of the numbers, Barry, and today in an hour I bet we’ll have something interesting to say…

    Permit me to post this while we wait:

    http://www.latimes.com/news/local/la-me-california-budget2-2009jul02,0,4522104.story

    “Chiang was set to print 28,742 IOUs starting at 2 p.m., said Garin Casaleggio, a spokesman for the controller. The initial warrants, which total $53.3 million, will go primarily to people who are expecting state income-tax refunds. The state last issued IOUs in 1992.”

    ,,,,I couldn’t help but be struck by the fact that the first IOU’s won’t go to any entity with clout…instead it will go to taxpayers expecting refunds for taxes paid….this is just about what you’d expect in this time of cynicism for our governmental leaders…it can’t be a good thing..

  48. leftback says:

    Housing prices appear to be falling in Bankhattan.
    http://www.bloomberg.com/apps/news?pid=20601213&sid=aKJTtME9cUhY

    The credit market still seems to be overlooking the CA funding crisis almost entirely.
    It’s the world’s 8th largest economy for chrissakes !!

  49. Onlooker from Troy says:

    I am not a lawyer of any type, and maybe it’s been addressed here lately and I’ve missed it, but how is CA’s issuing of IOUs not the minting of money? And therefore unconstitutional. I’m going to have to do some looking around to see what I can find on this.

  50. batmando says:

    Steve Barry at 9:51 am
    Good point…a business can cut a worker or save the same money by not firing someone, just cutting everyone’s hours.
    Case-in-Point:
    As I reported last month, 26 FT employees here at SW Ohio mfr of heavy-duty ag products processing machinery, cut back to 4 day weeks in June, going to 3 day weeks in July if no new orders are confirmed this week. Come August, if no new orders, lay-offs commence.
    4-day weeks for 26 employees = ~5 employees
    3-day weeks for 26 employees = ~10 employees
    so at least 10 jobs will be gone come August if no new orders are confirme