Coming Soon: Big Birth Death Adjustments?

When it comes to the Non Farm Payroll data, I have been taking "the Under" for several years now. The bet has been "money" in the bank month after month.

This is the month I take "the Over." And likely next monthy too. In fact, I may take the over for the next 4 NFP in a row.

Not because the job market is so great — to the contrary, its been mediocre, and is running at a sub-par basis over the past 1,3 and 5 years. While economists have been comparing the present cycle with the 1994/95 cycle, I find that comparison wanting. Northern Trust’s Asha Banglore observes that "Payroll employment has grown between 1.2% and 1.6% on a year-to-year basis since the Fed started raising the federal funds rate in June 2004. This compares with payroll employment growing between 2.5% and 3.5% in the twelve months ended February 1995."

Hardly a house afire. What’s worse is that the growth sectors are primarily lower paying service industries (Food service, retail, hospitality). The prime exception has been construction, which has been hot Hot HOT! and pays much better wages.

The reason I am taking "the Over" is due to accounting. You see, starting in April, we get the prime months when the BLS performs that sleight of hand magic called the Birth Death adjustment. These few months seem to be the major upwards estimates (*January and July are the negative months).

Where aas Non Farm Payroll is a specifically measured count of employees at FICA paying/reporting firms, it represents a true measured form employment data. (Compared to Household survey, which is a softer data source — they "ask people" about their work, rather than actually recording hard tax data).

The relatively clean NFP data collection moves a bit towards the theoretical with this B/D adjustment. The methodology of the Birth Death Adjustment is a hypothesis, a best guess. "We think that the economy created this many new firms, and those firms which are so new they are not yet in our database, created this many new jobs."

Since we are big into the guesstimates this month, I’ll make my own:
I guesstimate that about 100k new jobs will be added to the NFP report
each month due to this ledger domain.
Add to that the Katrina relocatees, who finally started showing up in the
data last month, and you have a recipe for a hot couple of NFP
numbers. The B/D adjustment gets figured into the total numbers, not
just the monthly — but most of economics takes place at the margins,
and this can have a significant impact.

Hence, the Over.

The New York Post’s John Crudele states it thusly:

"What is happening is much simpler (and, in a way, more inister): starting
with the March number to be released Friday, the government starts making
optimistic assumptions about hiring spurts caused by new companies coming into

The government will assume that around 200,000 jobs a month are being created
by new companies in the April, May and June report.

If the next four jobs reports reflect the seasonal abnormalities that I think
they will, then the markets will be soon expecting a whole lot more rate hikes."


Crudele expects the Fed to buy into the numbers, and keep raising rates. I suspect they may be more heavily weighting the wage growth, rather than total employment numbers — and will keep raising rates.

A look at actual job growth on a year over year basis, however, reflects that things are somewhat less rosy than the BLS data may imply:

All Employees: Total Nonfarm
% Change – Year to Year NSA, Thous


Source: Northern Trust


While the growth is better than 2001-2003 (pre-1% Fed rates), it is still much weaker than the 1994/95 period.

Unfortunately, 1.3% growth fails to keep up with the nations population growth. In other words, as a percentage less people have jobs  relative to the total population. Note that some of these missing jobs are also unregistered illegals — a guest worker program will result in a readjustment of this data. 

Bottom lineI expect a strong NFP number, and that should make people believe the Fed will keep tightening. These days, the market is juts as likely to shrug it off.


*Its not all upwards revisions: In January and July, the Bureau presumes
that firms that could not make it are closing their doors; The past 3
Januarys saw Preliminary Estimates of B/D adjustments of Jan 2005 -280,
Jan 2004 -321, and Jan 2003 -211 (Note that Q2-Q4 get revised)


Now vs. 1994/1995 – Part I
Asha Banglore
Northern Trust Global Economic Research, April 04, 2006

Historical Net Birth/Death Adjustments

NYPost, April 4, 2006

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. thecynic commented on Apr 6

    the markets are clearly trading like they expect the Fed to cease the tightening campaign at the May meeting. yesterday the nasdaq 100 led the markets higher on a 10% gain in AAPL which popped off its year low (short squeeze?). if you looked at the S&P sectors, energy and materials were the best performers and gold is in striking distance of $600/oz. this will be an interesting year. the markets (hedge funds) are playing chicken with Bernanke, daring him to stop tightening.
    ECB didnt’ tighten this morning and the Euro got spanked.
    tomorrow will be a big piece of the puzzle. what will the 10YR note yield and dollar do on the print?
    like you say, stocks could interpret either way the recent price action in stocks clearly suggests a bet on multiple expansion on an end to tightening. a sell off in the long end of the curve could spoil that bet. a rally could give it support.

  2. pauvre commented on Apr 6

    Interesting that the birth-death adjustment looks seasonal. My understanding is it’s an extrapolation (by ARIMA) that does a good job of removing any patterns from the residual errors. Never found anything on the model parameters, only on the general method. It almost sounds like an artifact of business tax reporting — the births, at least.

  3. cm commented on Apr 6

    My understanding is that BD “adjustment” only affects the preliminary numbers, and the final numbers are based on actually reported counts. (Hence sometimes substantial restatements.) Can anybody confirm/correct? — Thanks.

  4. calmo commented on Apr 6

    The methodology of the Birth Death Adjustment is a hypothesis, a best guess.

    And the guessin is pretty good as evidenced by the smaller annual baseline adjustments.

    But I’m with you (and those NT economists)– the Fed needs any excuse to keep raising the rates and lofty employment numbers , (no matter how many of them are for house cleaning), is such an easy, (unmissable even) [ie Cheney proof], target.
    Undoubtably, we have rampant latent inflation just waiting to jump out of the closet.

    This is huge: (Just when we were about to get over Katrina/Rita which was huge, this is huger.)

    Note that some of these missing jobs are also unregistered illegals — a guest worker program will result in a readjustment of this data.

    Can we trust the 11.7M number of illegal aliens? Doesn’t this issue make a mockery of all the unemployment stats if we don’t even know the actual number?

  5. Bear Mountain Bull commented on Apr 6

    Strong Jobs Report – Good News or Bad News?

    As always, there is a lot of discussion surrounding tomorrow mornings monthly non-farm payroll report. For quite some time, pundits had been complaining about the jobless recovery. But the jobs numbers have been a little better t…

  6. Zephyr commented on Apr 7

    Job growth of 1.3% annually is nearly a perfect match to the annual population growth of the US. (just over 1.2%).

  7. cm commented on Apr 7

    calmo: From my vantage point inflation is not very “latent”. I’m even getting unsolicited remarks from acquaintances to the effect.

    To elaborate (my specialty …), there are two sides to the coin — price hikes, and reduction in service/goods quality. Both quite visible to the trained (i.e. cynical) eye.

  8. todd commented on Apr 7

    Following business news in the NY Post?! Let me check USA Today and see what they’re saying first… LOL

    I keed… I keed.

  9. Lord commented on Apr 7

    On the longer term, boomers begin retiring in 2008, but insofar as I recall, the number of boomers is really quite modest for at least the first 5-7 years. Are people really expecting a far larger change sooner than will really arrive?

Posted Under