A standard rule of thumb is net ~125k jobs per month just to keep up with population growth. The economy added 114K net in September, which was only slightly weaker than prior months. A comment left by VitaminV on the story by Bloomberg.com says:

Why is there such a huge discrepency [sic] between household survey jobs added (873k) and establishment survey jobs added (114k) aka nonfarm payrolls? If you look at the data tables on the Bureau of Labor Statistics, its pretty clear 873k of new jobs in household data was pulled out of thin air (BLS attributes increase to self employed and part time job increases). Number of self employed is easy to fudge.

As for part time jobs – if there was really such a huge increase, then
a) why weren’t they registered in the establishment survey?
b) why did average hours worked per week did not actually decrease but increased instead to 34.5 hrs?

have a look at the data on the Bureau of Labor Statistics. its pretty straightforward and clear.

Household survey data:
http://www.bls.gov/news.releas…

Establishment survey data:
http://www.bls.gov/news.releas…

If you look at the trend in nonfarm payroll additions, this month is the weakest it has been for the past few months:

Sept ’11: 202k,
Jul ’12: 181k,
Aug ’12: 142k,
Sep ’12: 114k

I think Obama and co. had no choice but to cook the household data.

I haven’t looked at the data mentioned in this comment. I don’t think I need to look at that specifically because the facts about the recent condition of the employment market just don’t add up to a 3/10 reduction in unemployment to 7.8% considering the information stated above and little change in the number of jobless claims which have been around 370,000 per week since April.

As for the question about hours worked, I can see how the hours worked would increase if more people were working who did not have a job August, even part time or in self-employment. If they fudge one number, it makes sense to fudge that one was well as it would look strange if it wasn’t.

From the perspective of Market Monetarism and QE3, it is possible that unemployment might have responded to monetary easing. But I don’t know that much about the dynamics of the expectations channel and how long it takes to filter down into the unemployment numbers. It is a lagging indicator and the markets responded to the rumor of QE3, so it was already most of the way priced in by the time it was announced. To fully understand this, I will have to wait for some MM experts to weigh in on the jobs report.

Update: The links in the comment didn’t come over so I thought I would post what VV was talking about

Year

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

2002

-346

737

-261

-51

413

-124

-2

292

597

-294

-487

-95

2003

991(1)

65(1)

-48

199

-89

246

-316

75

60

375

440

-13

2004

61(1)

70(1)

-89

227

172

322

382

17

-86

245

499

-106

2005

120(1)

140(1)

269

600

355

105

312

408

-33

147

-49

253

2006

398(1)

307(1)

284

20

328

264

-151

423

190

499

220

436

2007

58(1)

29(1)

263

-734

317

160

-158

-223

562

-298

649

-322

2008

124(1)

-240(1)

-49

22

-201

-191

-208

-334

-137

-267

-714

-750

2009

-1141(1)

-527(1)

-906

-100

-360

-291

-112

-433

-683

-374

206

-639

2010

532(1)

165(1)

171

470

34

-203

2

199

6

-272

-135

283

2011

110(1)

221(1)

213

-136

180

-423

65

304

353

190

317

176

2012

847(1)

428(1)

-31

-169

422

128

-195

-119

873

1 : Data affected by changes in population controls.

Update 2

I took the links out of the comment because they were broken anyway. I copied the comment from Disqus and so it was broken from the beginning. Sorry about that. But the link I have in my first update works.

Regarding the data, the 873 number looks very odd. The only other time we have had that many jobs added in one month was in January of this year and January 2003, and it appears the there’s some statistical adjustments for that month that cause a bit of overstatement. Except for 2002, September hasn’t generally been a spectacular month for job creation. And while the household survey is useful for picking up employees at newer firms, either full or part time, I think the number for September 2012 is wrong and will be reversed at some point because there aren’t that many new firms that wouldn’t be participating in the establishment reporting. The establishment survey is more in line with what we know about the labor market which means the unemployment rate likely didn’t change.

I can’t say whether there was an effort to make the number look better than it actually is, and so I have to reserve judgement while taking this month’s jobs report with a grain of salt.

Advertisements