Question About The Jobs Report - Household Survey, The 680,000 Jobs

[Yavin4]

The Republicans are saying that the household survery is showing a net increase of 680,000 jobs in July, but the payroll survey from the corporations is only showing 32,000 jobs created in July. So, why the discrepancy? Here's my theory. Please tell me if it's accurate or not:

My Theory:

Because the job market has been miserable for four years running, people are more likely to take real short term temp work, even if it's a day or two. So, let's say Jane Doe takes a three day temp assignment. She won't show up on the payroll of the company that used her services, and she may not show up on the payroll of the temp company. However, when she gets called by the BLS, she shows up as having a job because if you work during a two or three week window around the phone survey, you are counted as employed.

[trotsky]

1. Here's my theory.

They'll cherry-pick whatever the hell they want. Supposedly the very same household survey showed a net LOSS of a quarter million jobs back in February. Didn't hear them bringing that up then, did ya?

But your theory sounds reasonable too.

[HuckleB]

2. Check out Paul Krugman's explanation in the NY Times from Tues., Aug. 10th

Beautifully explains this nonsense.

[faithnotgreed]

5. is it possible you can provide a link.. thank you

[ProfessorGAC]

3. Not Bad At All

Your theory is sound and is ONE of the causes. Another is that the household survey incorporates a much smaller sample size for the inference, so the 95%CI is MUCH, MUCH, wider, and there is, obviously a 2.5% chance that the number is really higher than even the prediction range.

The payroll survey is based upon a far larger sample size and is, therefore, more accurate on a single point basis.

So, couple this with the "short term temp job theory" (which we'll call the Yavin effect), and the disconnect becomes increasingly great.

Me, i'm more interested in the macroeconomic causes for the decline in jobs and the lack of job growth in a consumption driven GCP. The root causes are critical to me.

The Professor

[Zynx]

4. Simple: The Household Survey is useless.

It only surveys about 30,000 households and extrapolates those numbers across the entire economy. The Corporate(or Establishment) Survey surveys over 350,000 BUSINESSES meaning a much larger sample. I don't need to explain any more than that.

[Yavin4]

6. Thanks

That explains everything.