Image Source: Pixabay

Last month the BLS padded unadjusted jobs by 412,000. Curiously, not even the BLS can tell us how that impacted the headline-reported job gains.

Understanding the Birth-Death Model

Every month, for 10 years, I added this caution in my jobs report

Birth Death Model

Starting January 2014, I dropped the Birth/Death Model charts from this report.

The birth-death model pertains to the birth and death of corporations not individuals except by implication.

For those who follow the numbers, I retain this caution: Do not subtract the reported Birth-Death number from the reported headline number. That approach is statistically invalid.

The model is wrong at economic turning points and is also heavily revised and thus essentially useless.

I never explained in detail my caution. Starting next month, I will link to this post.

Here are some key details from the BLS link above. The third paragraph is likely to spin your head.

There is an unavoidable lag between an establishment opening for business and its appearance on the sample frame making it available for sampling. Because new firm births generate a portion of employment growth each month, non-sampling methods must be used to estimate this growth.

Earlier research indicated that while both the business birth and death portions of total employment are generally significant, the net contribution is relatively small and stable.

The second component is an auto-regressive integrated moving average (ARIMA) time series model designed to estimate the residual birth-death employment not accounted for by the imputation. The historical time series used to create and test the ARIMA model was derived from the UI universe micro level database, and reflects the actual residual of births and deaths over the past 5 years.

The net birth-death model components are unique to each month and exhibit a seasonal pattern that can result in negative adjustments in some months. These models do not attempt to correct for any other potential error sources in the CES estimates such as sampling error or design limitations. Note that the net birth-death forecasts are not seasonally adjusted, and are applied to the not seasonally adjusted monthly employment estimates to derive the final CES employment estimates.

The third paragraph is likely to confuse anyone. But the key to understanding the insignificance is in the last paragraph above, especially the last sentence.

Seasonally Adjusted vs Unadjusted Nonfarm Jobs

Last month, the BLS reported 157,984,000 nonfarm jobs. Of that total, 412,000 were due to the Birth-Death Adjustment.

Thus, the Birth-Death adjustment padded unadjusted nonfarm jobs by ((412,000/157,984,000) * 100) which is a tint 0.26 percent.

The BLS then took the 157,984,000 jobs and seasonally adjusted the total to 156,888,000.

The seasonal adjustment subtracted 1,096,000 (over a million jobs) from the unadjusted number despite the 412,000 birth-death addition, but that is an invalid comparison.

Note the seasonally adjusted number rose from 156,738,000 to 156,888,000 despite the huge decline in the unadjusted number.

The resultant headline was “Jobs increase by 150,000.”

Lovely! But how did the 412,000 impact the seasonally-adjusted gain of 150,000?

My Conversation With the BLS

• Mish: How much did the Birth-Death adjustment impact the seasonally-adjusted gain?
• BLS: Because of our methodology, we don’t know.
• Mish: What is the seasonally-adjusted Birth-Death number?
• BLS: Because of our methodology, we don’t know.

That conversation was not from last month but rather from about 10 years ago. Also. the conversion above is condensed. The BLS was polite, not abrupt as presented above.

The following process, derived from a long conversation, explains why the BLS itself does not know how the birth-death number impacts the headline number.

Five-Step Process

1. The BLS calculates the unadjusted birth-death number.
2. It applies that adjustment to its unadjusted overall estimate for the month.
3. Then it takes the unadjusted total and seasonally adjusts it.
4. Then it compares the seasonally-adjusted number for this month and subtracts the seasonally-adjusted number from last month.
5. The difference between the seasonally adjusted numbers is the headline total.

That’s why the BLS cannot say, nor can anyone else say, how much the birth-death model impacted the seasonally adjusted headline number.

Here’s another important idea I have been repeating since January.

Of the 894,000 rise in employment in January, 810,000 was due to annual benchmark revisions. And the BLS does not say what months were revised, just poof, here you go. 

We cannot, with strong confidence, suggest these reports portray an accurate picture of either jobs or employment.

Discrepancy Between Nonfarm Payrolls and Employment Levels

For a look in detail, at the latest jobs report including a discussion of the huge discrepancy between jobs and employment, please see BLS Jobs Report Stronger than ADP, Fueled in Part by End of UAW Strike

More By This Author:

A Curious Claim That The BLS Is Overstating Rent And Exaggerating Inflation
The Jobs Boom Is Clearly Behind Us, So What’s Ahead?
A Big Decline In Quits Suggests The Labor Market Is Back To Normal