Image Source: Pexels

The BLS BED report provides further confirmation the BLS Birth/Death jobs model is seriously screwed up.

Data from BLS, chart by Mish

Please consider the very lagging BLS Business Employment Dynamics report for the second quarter of 2022.

The BED report analyzes net job creation from new business and firms going out of business.

Key BED Statistics

• From March 2024 to June 2024 private-sector gains were 7.6 million, a decrease of 17,000 jobs from the previous quarter.
• Private-sector losses were 7.8 million, an increase of 639,000 jobs from the previous quarter.
• The difference between the number of gross job gains and the number of gross job losses yielded a net employment decline of 163,000 jobs in the private sector during the second quarter of 2024.

BED Statistics by Firm Size

• In the second quarter of 2024, firms with 1 to 49 employees had a net employment loss of 259,000.
• Firms with 50 to 249 employees had a net employment gain of 57,000.
• Firms with 250 or more employees had a net employment decline of 25,000.

Small businesses are struggling like mad. This is something the ADP reports also show.

Establishment Births and Deaths

• In the second quarter of 2024, the number of establishment births (a subset of the openings data) increased by 4,000 to a total of 326,000 establishments. These new establishments accounted for 975,000 jobs, an increase of 41,000 jobs from the previous quarter.
• Data for establishment deaths (a subset of the closings data) are available through the third quarter of 2023, when 927,000 jobs were lost at 295,000 establishments, an increase of 40,000 jobs from the second quarter of 2023.

Although the BED data on net jobs and the number of new businesses is through 2024 Q2, the number of closing businesses is only through 2023 Q3.

However, the important number is employment and that is through June of 2024.

Birth-Death Model

The lead chart shows that it is very rare for there to be net losses due to the birth and death of businesses.

Prior to covid, the trend of business openings and closings were similar. Now they wildly differ.

This is important because the BLS monthly jobs report includes a factor called the Birth-Death Model in which the BLS estimates the numbers of jobs created by the birth and death of businesses.

I have many times stated the monthly jobs reports are bogus due to sampling errors and birth-death assumptions.

BLS Survey Response Rates

Survey Response Rates from the BLS, Chart by Mish

Response Rate Notes

• CES: BLS Current Employment Statistics – The Monthly Nonfarm Payroll Establishment Survey
• JOLTS: BLS Job Openings and Labor Turnover Survey

The pathetic response rate for nonfarm payrolls is 43.3 percent down from 60.0 percent pre-Covid

The pathetic response rate for JOLTS is 33.4 percent down from 58.4 percent pre-Covid.

Response Rates and Survival Bias Issues

• First, firms that go out of business don’t respond. This is known as survival bias.
• Second, larger businesses are more likely to have someone who routinely answers such questions.
• Third, businesses of all sizes are deciding they have better things to do than answer BLS surveys.
• Fourth, businesses that are doing well are more likely to spend the time than businesses that are struggling or understaffed

BLS Sample Sizes

• Quarterly Census of Employment and Wages (QCEW) : 97 percent of bobs
• Business Employment Dynamics (BEDS): 92 percent of jobs
• Current Employment Statistics (CES): 27 Percent

The above numbers from CES FAQ

Of that 27 percent, the response rate is 43 percent with serious survival bias issues detailed above.

BED Net vs Nonfarm Payrolls

BED vs Nonfarm Payroll Quarterly Change

As noted, nonfarm payrolls (CES) represent all jobs whereas BED is only net business creation.

However, the relationship between BED and CES was stable pre-covid and now is anything but stable.

For five straight quarters, nonfarm payrolls wildly exceeded net BEDS jobs.

Birth-Death Model Jobs

Does anyone believe we added 653,000 jobs in the second quarter of 2024 due to net business creation?

Before you answer, please note those are unadjusted numbers. The BLS takes those unadjusted numbers adds them to unadjusted CES numbers then seasonally adjusts the mess for its monthly jobs report.

Q: What is the seasonally-adjusted birth-death number?
A: I don’t know, you don’t know, and not even the BLS knows.

Seriously, the BLS itself has no idea and that is what they have told me on multiple occasions.

It is wrong to subtract unadjusted numbers from smoothed and massaged adjusted numbers.

The Birth-Death numbers for 2024 Q2 were 363,000 in April, 231,000 in May, and 59,000 in June making the quarterly total 653,000 jobs.

653,000 jobs seems like a lot, and it is, but relative to total nonfarm payrolls of nearly 160 million its not that great.

December 2024 Example

• Unadjusted Nonfarm Payrolls: 160,458,000 by flawed survey
• Unadjusted Birth-Death by flawed estimate: -52,000
• Reported Seasonally Adjusted Number: 159,536,000 up by 256,000

I tried working with unadjusted birth-death numbers vs unadjusted BEDS but got similar wild-looking results.

BEDs vs Birth-Death Unadjusted Numbers

• 2024 Q1 Birth-Death: +9,000
• 2024 Q1 BEDS: -1,522
• 2024 Q2 Birth-Death: +653,000
• 2024 Q2 BEDS: -163,000

Seasonally-Adjusted BEDs vs Unadjusted BEDS

• 2024 Q1 BEDS SA: +493,000
• 2024 Q1 BEDS NSA: -1,522,000
• 2024 Q2 BEDS SA: -163,000
• 2024 Q2 BEDS NSA: +2,355,000

Four Things We Know

1. QCEW and BEDS confirm the nonfarm payrolls are substantially overstated and have been for about two years.
2. The BLS Birth-Death model assumptions are very flawed vs the hard data of BEDS and QCEW.
3. The Fed and economists are using highly unreliable and overstated jobs data to make interest rate decisions.
4. The BLS needs a serious overhaul of its data collection and modeling methods.

I hope you appreciate just how messed up this is.

I believe that why we have such a discrepancy between Gross Domestic Product (GDP) and Gross Domestic Income (GDI), numbers that should match but don’t.

Related Posts

January 30, 2025: 4th-Quarter Real GDP Increased 2.3 Percent, Good, Bad, and Ugly Details

The BEA reports GDP rose 2.3 percent vs a consensus 2.6 percent. But details are mixed.

Unfortunately, we will not have GDI numbers for two more months and it will be 5 months before we have a revised 2024 Q4 GDI.

September 5, 2024: Small Businesses Reducing Workers for the Last Four Months

ADP data shows small businesses with 1-49 workers have been reducing workers for four months. Those with 20-49 workers have shed workers for 7 straight months.

It’s been a while since I updated my ADP charts and will do so in February. But note that CES confirms ADP. In the second quarter of 2024, firms with 1 to 49 employees had a net employment loss of 259,000.

Small businesses are struggling.

And among other errors, it’s highly likely the BLS is oversampling large businesses or getting whacked hard on survival bias in its surveys.

More By This Author: