The Accelerating BLS Birth/Death Adjustment

I’ve mentioned the B/D adjustment over the years, and how its become an increasingly large portion of the reported BLS new jobs.

What I haven’t previously mentioned is that over the past year, it is accelerating: the Birth/Death Adjustment has become an ever-large portion of the reported NFP payrolls.

How much larger? Well, consider the following data points: Over the five month period ending in June, BLS B/D adds was a total of 922,000 new jobs. During the same period, the actually head counted Non-farm Payrolls (NFP) job creation was 709,000.

That’s right, fictional Birth/Death job adds have been outpacing actually measured job creation by some 30%.

As they do every year, BLS Net Business Birth/Death Model deleted jobs in January — in 2007, it was 175k.  That means that year-to-date, the net fabricated BLS new jobs was 747k — versus NFP growth of 871k — that’s 85.58% of NFP job growth.

Example of the absurdity of the new Birth/Death model — in place since 2001 — can be found in the specific employment sub-sectors.  Construction jobs are an obvious error (housebuilders added 12,000 workers), big jumps in education while school is out for summer is another, ‘Leisure & Hospitality’ B/D jobs are a multiple of the net category jobs created.

Prior to 2001, the B/D adds were less than 20k per month. Now, they dominate the Non Farm Payroll report.

Beneath the headlines, we see a far different reality. The FT noted:

“Wall Street economists, meanwhile, were surprised by continued hiring in the construction sector seen in Friday’s figures as  despite a prolonged housing market slump…The bulk of the hiring was in the service industries, as employers such as banks, insurance companies, restaurants, added 135,000 workers last month after hiring 199,000 workers in May. But they also pointed to signs of potential economic weakness, as the retail sector cut 24,000 positions."

Economists also said an unwelcome percentage of last month’s hiring was attributable to state and local governments, which added 40,000 staff and are not viewed as good indicators of economic activity."

I mentioned my incredulity over the fawning WSJ page 1 headline this past weekend (How Good Was NFP Really?). That is the soft prejudice of low expectations in action.

My favorite skeptic on BLS data is Bill King; Bill is even more incredulous over the reaction to what was by all measures a mediocre jobs report:

"And once again, The Street and their fin media stooges bray about how bullish 132,000 NFP jobs are (CSFB’s chief economist called the report ‘excellent.’) even though just a few years ago Street Conventional Wisdom held that the economy must generate about 175,000 jobs each month just to absorb demographic growth.

Lehman’s economist said, “The labor market is one of the stronger parts of the economy right now.” This is a Clintonesque, qualified statement. It could imply that the rest of the economy really sucks."

Sucks, indeed.

Here’s Bills’s chart of the past few decades NFP growth:

NFP with 1-year (12 month), 10-year and 40 year (480-month) moving averagesNfp_20_years

Source: M.Ramsey King Securities


As the above chart clearly shows, the June NFP number is hardly ‘excellent’ or indicative of economic strength. With NFP yearly average of 167,333, how can a release ~35,000 or > 20% less than the yearly average be ‘excellent’?

The 40-year average NFP growth is 150,600 — and that is with a national population considerably less than 300 million people. Anyone asserting a NFP number ~15% below the 40-year average as ‘excellent’ or a sign of strength is a shill who has failed to do the math . . .

UPDATE 3 July 11, 2007 8:20am

Here this morning’s article

Are Business Births as Rosy-Cheeked as BLS Says?

Caroline Baum

Bloomberg, July 11 2007

UPDATE 2 July 10, 2007 3:02pm

Here is Paul Kasriel’s take:



UPDATE July 10, 2007 1:20pm

Bloomberg’s Caroline Baum, who has a column on the same subject  tomorrow, points us to this explanation of the B/D adjustment from Kirk Mueller, BLS’ supervisory statistician:

Impact of business births and deaths in the payroll survey

June jobs growth higher than expected
Eoin Callan
FT, July 6 2007 14:30 | Last updated: July 6 2007 16:27

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. Sailorman commented on Jul 10

    It seems to me that the BLS like most government organizations (think FED) and corporate boards is attempting to manage the stock market, not do their proper jobs.

    Home Depot announced a stock buyback today. Does that make any long term sense for this company? It seems to me that Depot would be better served by using the money to improve the condition of their stores and their service. Short term stock price trumps sound business decisions.

    While the impact on the market is positive in the short term, this practice is becoming more widespread and will have a serious long term negative impact on the economy.

  2. blackvegetable commented on Jul 10

    The evidence that the “Cinderella Economy” is grossly over-rated is everywhere….

    here is more…

    the annual growth rate in real disposable income

    The Cinderella Economy – 2.44%

    The Malaise Years of Jimmy Carter – 3.11%

    The most inept stewards of the economy in the post-Depression era…..

  3. Francois commented on Jul 10

    In the 80’s, there was a say about the auto industry: “When confronted with a problem, Toyota, Nissan and Honda call their engineers. GM, Chrysler and Ford call in their lawyers.”

    Look who’s ahead in the game now.

    Likewise, whenever confronted with a problem, the financial industry in this country call in their spin doctors, paid shills and deal makers.

    How long can this continue?


  4. Michael Donnelly commented on Jul 10

    I guess its up to me to defend the BLS. First, they don’t give a rats a_s about the stock market.

    Does anyone here recall what happened in the last benchmark revision of jobs? BLS had to add nearly 1,000,000 jobs over the course of the year. In other words the B/D model should have been adding 80,000 more jobs per month than what it was adding.

    The main problem with the B/D model is it’s isn’t cyclical in nature, so when the economy is rising it undercounts, and when the economy is falling it overcounts.

    Has the construction sector added 153,000 new jobs (B/D count for Feb-June) ? No. Not a chance. BLS needs to add a forecast component to the B/D model, but not scrap it entirely.

  5. michael schumacher commented on Jul 10

    the home depot thing is outrageous. Goldman “upgrades” them based on nothing more than what if’s and then they trim guidance. Gotta think someone at GS used yesterday to exit. Let’s invent earnings and profitability via stock buybacks. Wonderful way to appease shareholders.

    I guess the mountain of consumer credit released yesterday means absolutley nothing as well. These experts forecasted a .2% rise and we got a 50% rise. I see it mentioned no where other than a feed on Forbes. I expected BR to have something on it this morning…….as I did send him the original story.

    All that nonsense over on HD and it’s up for the day already…sheer lunacy.


  6. zell commented on Jul 10

    Soon the gov. will be including an adjustment for illegal workers…..Anywho we are flying blind anyway. Yesterday isn’t where we are going.

  7. Joe Klein’s conscience commented on Jul 10

    As I said last week, if the leaders of this country think it is okay to lie us into a war, gaming the employment figures is no big deal.

  8. Fred commented on Jul 10

    All of the “employment numbers are bogus” analysis is interesting, and may well be the “truth. Yet one still has to ask why there are major shortages of employees in many situations.

  9. Jay Weinstein commented on Jul 10

    I offer the following topic for discussion:

    While there is no doubt that the consumer is struggling, the bulls do have merit in their argument that industrial non-auto America is booming. I can attest to that, having just had one of the best quarters in my firm history by avoiding retailers, home builders, the toxic financials, and owning small industrial widget makers. Their backlogs and prospects are as good as I have seen them, and prices have risen accordingly.

    I am totally NOT a technical guy, but if you look at the IBD 100, you will see an amazing number of these small to midsize industrials.

    So while I totally accept the bearish housing/mortgage camp, at what juncture does the consumer problem affect the booming sectors? Or does it at all? What might the mechanism/clues be?

    I am not interested in political bashing, or bitter bears who don’t understand why the market hasn’t fallen apart. Please try to keep opinions rational and informative.

    Thanks to Barry for the forum and a great site.

  10. Karl Smith commented on Jul 10

    To echo what Micheal Donelly said, there is nothing fundamentally wrong with an accelerating B/D adjustment.

    It results mechanically from the fact that the Birth/Death residual is in fact accelerating. That is, measured and imputed employment has consistently trailed benchmark estimates of employment.

    Now, not to nitpick but as I understand it the B/D model does have something of cyclical component, to the extent that it can be picked up in an ARIMA model.

    The problem is that the model does not adjust immediately for shocks. For example, the run-up in housing that we saw was no ordinary cyclical expansion. It was an unusual boom.

    The B/D model will interpret that as secular change, but it is in fact completely transitory.

    Long story short, yes construction is probably wrong, but that comes from the fact that construction has been so wild over the last few years not some deception on the part of the BLS.


    BR: Wrong. Its an absolute fabrication. Historically, this has been a minor adjustment — Now, its become the majority of job creation data. Thats juts wrong . . .

  11. michael schumacher commented on Jul 10


    Where is the wage pressure if everyone is gainfully employed and there are shortages of employees???

    If there was a tight labor market ( as the gov’t and you are suggesting) then salaries would reflect that…..I see no upward pressure on wages and neither does the gov’t because that would show inflation.

    You can’t have you’re fake statistics both ways.


  12. Greg0658 commented on Jul 10

    Jay – Is the defense and highway spending keeping things afloat here?

    As far as the President Carter slam (again) – he was a good man handed a load of crap – remember the gas prices, 3 Mile Island and Iran Hostage Affair. He’s still a good man.

    You think any President could have managed those items any better? That President Reagan swearing-in exchange was something. How many years has this agenda been brewing?

  13. Stuart commented on Jul 10

    BRAVO! BRAVO! but alas, to no avail. The shills and montebanks don’t care. They are past the point of commitment with their a spinful purpose.

  14. Frank de Libero commented on Jul 10

    I made a post yesterday about the NFP but as a story. It supports Barry’s continuing analysis on this subject, eg, “..just average or mediocre job growth would have produced 190,000 jobs in June instead of 132,000. That doesn’t fit with the happy talk about the June report, or previous reports.” For those interested, the post is here:

  15. will rahal commented on Jul 10

    I have shown that when the(3-month moving avg) Unemployment rate increases year-over-year, an accelerated decline in Wages takes place, leading to recession.
    So I was trying to come up with what monthly
    change in Payroll will lead to an increase in the unemployment rate. It is not straight forward because the Labor Force does not change at a constant rate.
    Nevertheless, I came up with 143k for the
    monthly payroll needed to keep (in this economic environment) the unemployment rate steady. See

  16. joe commented on Jul 10


    Not sure what your beef with the HD news is. People on this site rightly condemn share buybacks as in general value destroying. But, there are times when buybacks are the OPTIMAL use of capital. HD is a mature cash cow, generating capital far in excess of whats needed to run the business. In addition, it was capitalized as a growth company – meaning conservatively – not as a mature, slow growth business. Levering up the balance sheet is a natural progression in the life of a firm.

    I’ll leave the final word to Warren Buffett:

    “There is only one combination of facts that makes it advisable for a company to repurchase its shares: First, the company has available funds — cash plus sensible borrowing capacity — beyond the near-term needs of the business and, second, finds its stock selling in the market below its intrinsic value, conservatively-calculated.”

    These two criteria exactly match HD.


  17. michael schumacher commented on Jul 10


    I do not even know where to begin with that….

    mature Cash cow means that you do not go and borrow almost $23 Billion dollars to buy back shares while you’re stores are falling apart.

    Home depot is a unique example of financial engineering at it’s best (or worst) depending on how you look at it. One would think that you could put someone in there to grow the business organically however HD has been run like GE and not in a good way. Financial engineering is all these people know how to do. hence they can’t figure out how to increase profit margins so they just decrease the amount of stock available and presto the SP rises and then everyone is happy….for a short time.

    This is not a “conservative” strategy it’s desperation


  18. Aaron Byrnes commented on Jul 10

    Someone should normalize Bill Kings B/D chart to take into account the lower population in the past so we can better see how strong the recent NFP data really is.

  19. joe commented on Jul 10


    First of all, they’re borrowing 12.5B, the other 10B of the repurchase comes from cash on hand and the 9.5B in proceeds from Supply. But why quibble about a little matter of 10B.

    You ask why do a buyback when they need to reinvest in their stores? Guess what, they generate so much cash, they can do BOTH. And they are. They’re actually massively reinvesting in the existing stores, but again, why let facts get in the way of your argument.

    Again, in general I agree that buybacks are financial engineering, but in this case, it’s value creating, not destroying.

    It’s not desperation, it’s acting in the best interest of long term shareholders.

  20. michael schumacher commented on Jul 10

    totally disagree……

    It’s creating something that is not there, and will never be there. It’s only purpose is to prop up a share price via the only way they know how to… engineering.

    What ever happened to growth for growth’s sake???

    Not at HD when you can create it by decreasing the amount of shares and giving a short term pop. I guess it would be easier to stomach if they did’nt just take almost 20% away from guidance.

    It smacks of creating value when (by their own admission) it does’nt exist organically.

    By your own numbers they are only using $500 million of cash since HD supply actually was profitable and that was sold thus generating the cash to purchase shares and more importantly LOSING the income from that operation. That’s not really cash on hand since an asset was sold. Do I need to point that out to you????

    It is creation….plain and simple. They can’t cut it on there own and have resorted to engineering a way to show share price appreciation they certainly are not going to realize cash flow from HD supply so let’s just use that money to create it differently thus shifting the liability for doing it back squarely onto existing shareholders.

    Yea that’s value creating….but for whom???
    not the shareholders who are now saddled with debt for no other reason than greed by upper management.


  21. some guy commented on Jul 10

    I posted on this subject the a few days ago on the last NFP post, and if i recall no one even commented on it. You are all missing a very important point that is creating a flaw in your analysis (although in my opinion the conclusion doesn’t change).

    You are comparing seasonally adjusted headline pay roll numbers with non-seasonally adjusted birth/model numbers and coming up with an apples to oranges comparison. The numbers you should use to compare is the ACTUAL NFP payrolls data not the seasonally adjusted numbers.

    To summarize the number of B/D adjustments this year was 747k as you quoted, however the actual number of jobs added since Feb was 1168k. So net of the B/D model actual jobs added was 421k an average of 70k a month. So the average total monthly UNADJUSTED numbers so far this year have been 195k a month, B/D has averaged 125k a month or about 64% of the ACTUAL employment gains as reported in the NFP report.

    I don’t know how this relates to historical relationships between the actual reported numbers and the B/D model, but my intuition says that it shouldn’t be such a high number and is still a warning sign we should all be aware of.

  22. blackvegetable commented on Jul 10

    “As far as the President Carter slam (again) – he was a good man handed a load of crap – remember the gas prices, 3 Mile Island and Iran Hostage Affair. He’s still a good man.”

    My point was not to “slam” JEC (disclaimer: I didn’t vote for him at either opportunity)….for decades we have been told that he was, without question, the worst POTUS since Harding, specifically with respect to his economic record…..

    that is a lie…..and a gross dis-service to a good man, and at worst, an average president.

    I use the “Malaise Years” as a yardstick against which to measure the “Goldilocks Economy”…….You would be hard pressed to find a meaningful metric by which Junior bests Jimmy…..Skeptics are free to look at GDP and nominal job growth in order to test that assertion.

    The more insistent we are in pointing out the facts to the Herd, the more likely it will be that they will not be so easily snowed next time some Upper Class Twit sells the snake oil of Supply Side……

  23. michael schumacher commented on Jul 10

    and on cue Ber”hank”e utters the dreaded inflation word and the market reacts as if it were a surprise that they are spouting the same old tired line. “1-2%” yep…thanks for nothing……this is just like the surprise over at DHI this morning….



  24. rex commented on Jul 10

    “Some guy beat me to the punch” with his comment at 12;42.
    Barry: your analysis of the birth-death model is just wrong. You have to compare nonseasonally adjusted data to nonseasonally adjusted data.
    That’s not to say that your main point is wrong (that the economy sucks), it’s just that you made up facts to support your conclusion.
    Sort of like what you accuse the bulls of doing….

  25. VJ commented on Jul 10

    just a few years ago Street Conventional Wisdom held that the economy must generate about 175,000 jobs each month just to absorb demographic growth.

    During the ’90s, the threshold more frequently utilized was 185,000.


    As to those who believe the fictional B/D jobs are real, then explain the dichotomy of having:

    * Bankruptcies at the highest level in history

    * Home foreclosures at the highest level in history

    * Inventories of unsold homes at a 15 year high (the largest inventory in relation to sales since January 1991, when the national economy was in recession, under Poppy Bush)

    * Federal budget deficits at the highest level in history

    * The savings rate go negative for the first time since the Great Depression

    * The total amount of money Americans have to spend after taxes relative to overall output of goods and services (U.S. Disposable Personal Income as a percentage of GDP) hit the lowest level in 25 years

    * Poverty increase every year this administration has been in power (after declining every year of the previous eight years)

  26. some guy commented on Jul 10

    I did some research. Here is the percentage of the unadjusted employment number attributable to the Birth/Death Model for each year since 2003:

    2003 693.55% ** anomaly due to the low number of jobs that year
    2004 40.80%
    2005 34.06%
    2006 43.13%
    2007YTD 63.96%

    I also compared the Jan-June numbers given that’s where we are so far in 2007 to check to see if the lack of data for July-Dec for 2007 explained the 2007 anomaly and it did not:
    2003 -1037.21% as of June 2003 YTD job growth was negative and very small in absolute terms
    2004 40.41%
    2005 39.14%
    2006 46.73%
    2007 63.96%

    2007 employment seems very likely to be overstated by the birth/death model even using the correct numbers.

  27. rex commented on Jul 10

    The economy doesn’t need to create as many jobs as it used to. Certainly not 185,000 per month.

    For example, In 2006, the adult population (16-65) grew by 2.15 million or 179,000 per month, according to the Census Bureau. Figure that about 2/3 of them will join the workforce (the rest are in school, in prison, in the army, taking care of kids, disabled, etc). That leaves about 119,000 jobs to be added each month to absorb population growth.

    The economy needed to create more jobs in the past because: 1) the adult population was growing faster, and 2) the participation rate for women was growing.

  28. m3 commented on Jul 10

    rex, that job growth analysis of yours completely neglects the effects of immigration, among other things.

    i don’t buy that argument at all…

  29. Greg0658 commented on Jul 10

    I Wiki’d ‘Malaise Years’ and it came up with (way down list) this link

    After reading on those ressecion years I was saved by 3 Mile Island in a way. I spent alot of time in a local nuke retooling it. We (USA) pulled out of those days by tooling up China for WalMart. What we gonna pull out of asshat this time? And are the hedges the next savings and loan crisis?

  30. me commented on Jul 10

    “The economy doesn’t need to create as many jobs as it used to. Certainly not 185,000 per month.”

    So everybody is rich and we don’t need jobs? And where did you get the 185,000? Under Clinton for 8 years the average was over 225,000 per month. This is the worst economy of my lifetime.

  31. VJ commented on Jul 10


    The economy doesn’t need to create as many jobs as it used to. Certainly not 185,000 per month.

    I don’t know how to break this to you, but with an even larger population, the national economy requires even MORE jobs than it did in the ’90s just to keep even with the demographic growth.

    BTW, the “participation rate for women” saw it’s largest transformational shift in the 1980s under Reagan. Before Reagan entered the White House, the majority of married women did not work outside the home. After Reagan exited the White House, the majority of married women worked outside the home.

    They were forced to, as a direct result of the overwhelming majority of American’s real wages going backwards by about 20% during the 12 years of Reagan/Poppy Bush.

  32. rex commented on Jul 11

    to m3, me and VJ: It’d be nice to have some facts to back up your arguments. Then we could test them. Try it sometime. It’s almost as rewarding as saying “worst president ever.”
    The population growth figures I cited came from the Census Bureau. Your population growth figures came from… where? Oh yeah. You didn’t provide any.

  33. VJ commented on Jul 13


    The population growth figures I cited came from the Census Bureau.

    Unfortunately, that data comprises only half of an equation.

    Try the National Labor Participation Rate, or the Civilian Employment-to-Population Ratio:


  34. some guy commented on Jul 23

    I did some research. Here is the percentage of the unadjusted employment number attributable to the Birth/Death Model for each year since 2003:

    2003 693.55% ** anomaly due to the low number of jobs that year
    2004 40.80%
    2005 34.06%
    2006 43.13%
    2007YTD 63.96%

    I also compared the Jan-June numbers given that’s where we are so far in 2007 to check to see if the lack of data for July-Dec for 2007 explained the 2007 anomaly and it did not:
    2003 -1037.21% as of June 2003 YTD job growth was negative and very small in absolute terms
    2004 40.41%
    2005 39.14%
    2006 46.73%
    2007 63.96%

    2007 employment seems very likely to be overstated by the birth/death model even using the correct numbers.

Posted Under