Big difference btwn. non-adjusted vs. adjusted fig. on employment report

[Happy Hopping]

I have notice a huge difference between the Seasonally Adjusted figure on the Total Private sector figure vs. the non-adjusted Total Private sector fig.

For e.g., in Apr. 2010, the non-adjusted Total Private fig. job gain from previous month is actually 1,084,000

But on the report, the adjusted fig. for Total private gained from previous month is only 231,000.

How can they difference by that much? There should be a somewhat close relationship between the 2 figure knowing that they are seasonally adjusted

And having said that, 1.084 million job gain in a mth. is the real fig. for Private Sector, that's a great figure, isn't it? What did I miss?

 

[LeeD]

You are talking of total employment, not non-farm payrolls, right?

Seasonal jobs make a large difference.

 

[Happy Hopping]

I am definitely talking about Non-farm payroll, the job report on 1st Fri., of ea. mth.

http://www.bls.gov/news.release/archives/empsit_05072010.htm

Using Apr. as an e.g., I am talking about fig. like:

Total Private: 107143 for Apr. 2010, 106059 for Mar. 2010, as such the difference is 1,084,000 jobs gain.

 

[Happy Hopping]

Nobody? Because I thought when we look at Seasonally adjusted, we get the overall picture of the economy, but when look at the non-adjusted, say we look at Apr. to Jun. Non-adjusted job gained, we can tell a big gain in jobs in that 3 mth. for eg., whether it does /does not reflect on Q2 of certain co. such as retails, earnings.

For e.g., if we have 1M job gain per mth., non-adjusted, in Apr., May and June, we can understand certain retail co. Q2 earning should be great.

By contrast, w/ those high gain, if a certain retail co. still not doing well, it could be worse for certain retail co. if non-adjusted private sector turning worse.

it also show if a Q2 is good (or vice versa), it doesn't mean the next qtr. is good

 

[trefoil]

You're right as far as the effect on retail, and those unadjusted figures probably have something to do with why stuff like consumer confidence did a bit better than economists had expected.
What's annoying is that you can't get some figures, like hours worked, in unadjusted terms (unless someone out there knows a source??). Unadjusted hours worked would give you a better handle on whether gains would continue or not, since the tendency at the start of better times is to make folks work more hours, and only later actually hire new people. It would of course also clue you in on whether lots of folks are making good money working overtime.

 

[Martinghoul]

Non seasonally adjusted avg weekly hours etc are available on the BLS site here: http://www.bls.gov/sae/#data. Just use their data search tool and you shall find.

 

[trefoil]

That's state & metro area stuff, which I'd have to enter off the pdf, then sit & sum in order to get any idea of the nationwide value. Grr.
However, you did inspire me to try one more time to find something, and I did find that over here - http://www.bls.gov/ces/#tables - in table B-2 there's some unadjusted nationwide monthly data embedded if you look for it. It shows hours increasing to enough of an extent that a quick eyeballing seems to show weekly earnings for manufacturing employees went up by around 4-5% over the past year, which is pretty good info to have. I'll have to play around with this.

 

[Martinghoul]

True, that's only the state data with no national aggregation... However, you can get them all into a single .CSV file (by selecting "more formatting options" or something). You can then do your magic in Excel.