Drilling beneath the BLS Headlines

One of the fascinating things about the US Goverment’s data producers, including the BLS and the BEA is that they don’t seem to hide anything.

While conspiracy theories may be sexy, the reality is far more mundane. Its all there if you have the temerity to dig thru endless data (eternal vigilence and all that). This must be terribly disapppointing to the black helicopter/tinfoil hat crowd.

The amazing thing is that most people don’t bother. By "most people," I am referring to the economists, journalists, strategists and fund managers who trade off of this data.

So let’s do a little digging, and see if any nuggests of gold might be buried amongst the dirt:

• Once again, we see the Birth Death adjustment — thats the hedonic guesstimate which supposes the number of new jobs created by businesses so new they have yet to be measured — actually exceeded the number of new jobs. This month, its 184k. That’s about on par with last June’s B/D adj.

Note that this adjustment is not a one for one type — it goes into the BLS model further upstream in the massaging process. Regardless, its still a substantial number.

Household survey shows another 240,000 people left the Labor Force last month. That’s greater than either survey’s number of newly employed. The Civilian labor force participation rate actually decreased this past month. We still see unemployment going down because more people are dropping out of the labor than obtaining new jobs. That’s hardly cause for celebration.

U-6, the broadest measure of unemployment, actually ticked up to 9.0%. This measure includes:  Total unemployed, plus discouraged workers,
plus all other marginally attached workers, plus total employed part
time for economic reasons.

• Lastly, here’s an oddity I just noticed:  The jobless rates by race shows that Whites are no longer enjoy the lowest unemployment rate — that pleasure now belongs to Asians:

Asians: 4.0%
Whites 4.3%
Hispanics 5.8%
Blacks 10.3%

I don’t recall when that first occurred, but it would be interesting to track down. The chart on this (Household Data, Table A-2) only goes back to May 2005.

The headline data may appear benign, but digging beneath shows a less than robust economy. Not awful — just not as rosy as the scenario painted by the cheerleaders . . .

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. spencer commented on Jul 8

    Asian unemployment rate fell below the white rate ealy in 2004. It was also lower for a short while in late 2000-2001.

    The most interesting employment chart might be the
    monthly employment gains. They have been significantly lower this cycle then in previous cycles.

  2. spencer commented on Jul 8

    You may also want to look at the difussion index.
    It shows a much weaker picture.

  3. The Stalwart commented on Jul 8

    A Less than Rosy Picture

    Barry Ritholtz, today, continues his practice of puncturing the government’s economic data. Today it’s the BLS’ Non-Farm Payrolls report which on the surface seemed ok–195,000 new jobs. But digging through the fine print, there are clearly several fla…

  4. RT commented on Jul 8

    Once again, we see the Birth Death adjustment — thats the hedonic guesstimate which supposes the number of new jobs created by businesses so new they have yet to be measured — actually exceeded the number of new jobs. This month, its 184k. That’s about on par with last June’s B/D adj.

    There was a periodic discussion about the birth-death adjustment on Brad DeLong’s blog during much of 2004. DeLong’s opinion, which I respect, is that it’s fairly sound. I’d still love to see a study comparing its projections (for projections they are, even though they’re projecting a figure for a time period already past) with the first estimates that include/exclude the newly born/dead businesses.

    I don’t recall when that first occurred, but it would be interesting to track down. The chart on this (Household Data, Table A-2) only goes back to May 2005.

    Click your link, then click on “Get Detailed Statistics” in that blue band across the top of the page, then click on “Historical Data for the “A” Tables…” near the top of the page. Then check the boxes in the “not seasonally adjusted” column for white unemployment rate and Asian unemployment rate.

    For some reason, the seasonally unadjusted historical data on Asian unemployment is available back to Jan 2000, but the seasonally adjusted data isn’t available. Whatever: there’s no reason I can think of not to believe that seasonal adjustments act on all racial groups in similar ways. So aside from a blip in October 2004, it looks like Asians have had a lower unemployment rate than whites since the beginning of 2004.

  5. pgl commented on Jul 8

    You are right about the reason for the lower unemployment rate being simply the fall in the labor force participation rate – as the employment to population rate was unchanged. No cause for celebration indeed. But I predict a few of those NRO econopundits (cough, cough) will be celebrating by Monday. They always do.

  6. TimW commented on Jul 8

    Barry.. you are exactly right. The conspiracy theories regarding economic statistics always make me laugh… because I help to put together official gov’t statistics for my job. It’s a really unsexy job with really no chance of fudging the data.

    It is interesting to know the numbers and compare them to the market forecasts before the release though…

  7. Brad DeLong’s Website commented on Jul 12

    A Few Unemployment Details

    Barry Ritholz talks about unemployment: The Big Picture: Drilling beneath the BLS Headlines : One of the fascinating things about the US Goverment’s data producers… is that they don’t seem to hide anything…. Its all there if you have the temerity t…

Posted Under