Friday, November 5, 2010

Employment Reports Split

Marketwatch details:

Nonfarm payrolls rose by a greater-than-expected 151,000 last month as private-sector employers added 159,000 jobs, the Labor Department said Friday. The September number was revised to show payrolls fell by 41,000, less than an original estimate of a 95,000 decline.

However, the unemployment rate, which is obtained from a separate household survey, remained at a lofty 9.6% in October. About 14.8 million people who would like to work can't get a job. The jobless rate has been above 9.0% since May 2009, right before the recession ended.

Odd reports... the headline non-farm payroll number jumped more than expected.

But, the household survey disappointed, with the number employed actually falling and those leaving the workforce jumping by more than 400k.

Source: BLS


  1. Jake,

    I'm trying to figure it out, too. I'll let you know if I come up with something. It's weird.


  2. Here's what I came up with:

    As you know, the household survey showed there were 330,000 fewer people employed in October than September. But, this isn't a very accurate estimate, because the BLS only surveys 60,000 households. The difference could just be because of estimation errors. The 151,000 job gain from the establishment survey is based on 160,000 businesses, which is more statistically significant. Therefore, trust the employment numbers from the business survey.

    For more, see my blog post


  3. i get the larger sample size aspect, but why would the household sector (which in theory should catch the turns in the economy as a lot of jobs post-recession historically come from startups rather than corporate hirings) be lagging / reversing the trend...

    either way, i appreciate the follow up!