Random Thoughts on Today’s BLS Unemployment Report

This morning the Bureau of Labor Statistics has released its employment situation report for March 2012:

Nonfarm payroll employment rose by 120,000 in March, and the unemployment rate was little changed at 8.2 percent, the U.S. Bureau of Labor Statistics
reported today. Employment rose in manufacturing, food services and drinking places, and health care, but was down in retail trade.

  1. Expectations were for around 200,000. That’s far below expectations.
  2. The minor decrease in the unemployment rate suggests that people are continuing to leave the labor force (or not entering it).
  3. Those who said that the way the BLS does adjustments always causes the February report to look good appear to have been right.
  4. Those who said that the mild winter may have shifted some employment from March, April, and May into January and February appear to have been right.
  5. There will be some frenzied attempts to portray this as a good report. Will they focus on the unemployment rate?

We’d better hope that some of the swords dangling over our heads keep right on a’danglin’. The situation seems very precarious to me.

I’m not a bit surprised that drinking establishments have seen an up-tick in employment.

18 comments… add one
  • Icepick

    Okay, I heard about the report for the first time just now. I don’t even need to go looking for the particulars to know this one is BAAAAD. This is enough to barely keep employment steady, at best. A decrease in the participation rate is the ONLY way the UE-3 rate goes down in this circumstance.

  • Icepick

    Additional questions: How is the stock market flat on this news? Is it because it increases the chance of more Federal Reserve Bank System stimulus? The news that the Fed wasn’t going to help anymore sent stocks into the tank for two days, and this doesn’t even cause a ripple. Weird.

  • Icepick

    Additionally, it isn’t so much that more hiring would have taken place in January in February, as there would have been fewer job cuts. Only with Seasonal Adjustmens would there appear to be hiring. (If I have time, which seems unlikely, I’ll try to suss out the numbers later. But that is unlikely as once the wife gets home this evening I have too many things to do this weekend, and not enough time to do it.)

  • Drew

    Not only do points 1-5 make sense in an absolute sense, it’s hard to imagine the usual crew of commenters here disagreeing.

    As for point 4, at the risk if being criticized for “anecdotal” analysis, we just closed an acquisition that indeed has had a good Q1 based on mild weather. In fact, relative to last year, which had horrible weather, as the year over year months burned off profitability on a trailing twelve month basis increased by almost 25%.

  • Drew

    Ice pick

    Re: the stock market. The market clearly has reacted to artificially low interest rates, and corporate earnings, but in my opinion (and that of my financial advisor) gotten ahead of itself. With corporate earnings increases becoming increasingly difficult to keep pace with the last 18 months, and the market approaching “fair” value on a PE basis we may have a trading market through the election.

  • Additionally, it isn’t so much that more hiring would have taken place in January in February, as there would have been fewer job cuts.

    You bring up a good point. IMO those who’ve emphasized jobless claims have been whistling past a graveyard. Job cuts are a much more interesting metric.

  • Looking over the damage control reactions, the response seems to be “it’s just a blip in a generally improving trend”. My question would be “what trend?” to which I would add “business cycle”.

    Let me expand on that a bit. What I see is variation around a stable value of somewhere around 170,000. Sometimes above. Sometimes below. Cf. here.

    That’s not enough to bring the millions of unemployed back to work.

  • Ben Wolf

    One reason the stock market is relatively flat is that corporations are no longer swimming in profitability as they were at the height of the recession and height of the government’s budget deficits. A quick tabulation of the Kalecki Profits Equation using last year’s data shows that deficits were an enormous chunk of corporate profits. The deficit has fallen this year and will likely continue to do so unless the country slips back into recession.

    In a typical recovery this would not necessarily be a problem as consumer spending would expand faster than deficits would contract, but so far this doesn’t appear to be happening in 2012.

  • PD Shaw

    What I heard this morning from administration supporters is that people are leaving the workforce as a natural part of demographics — the baby boomers are retiring. That’s probably partly true, but my impression is youth unemployment is quite high.

    And if true, do other economic assumptions about future growth need to be revised?

  • michael reynolds

    A question I could probably answer myself, but hey, you guys aren’t doing anything. . .

    Does employment count the self-employed? Free-lancers, etc…?

  • Does employment count the self-employed? Free-lancers, etc…?

    Yes. But maybe not very well.

  • Ben Wolf

    @Michael Reynolds

    The BLS actually has a paper discussing self-employment:

    “Since January 1994, employed respondents in the
    monthly CPS have been asked the question “Last week,
    were you employed by government, by a private company,
    a nonprofit organization, or were you self-employed?”
    Respondents who say that they were employed by government,
    a private company, or a nonprofit organization
    are classified as wage and salary workers. Individuals who
    say that that they are self-employed are asked, “Is this
    business incorporated?” Respondents who say yes are the
    incorporated self-employed and are classified as wage
    and salary workers; respondents who say no are classified
    as unincorporated self-employed, the measure that
    typically appears in BLS publications. Since 1989, unpublished
    tabulations of the incorporated self-employed
    have been produced by the Bureau on a regular basis”


  • michael reynolds

    Thanks Dave and Ben.

  • Fewer people in the labor force, hence unemployment went down slightly:

  • michael reynolds

    So if a significant number of people became unincorporated self-employed you’d see the overall unemployment number drop as they took themselves out of the job market but also see the new jobs number drop, since, in effect they’d be doing a job?

  • Ben Wolf

    The unemployment number would drop, but depending on whether they catch and question a non-incorporated self-employed person the labor participation rate could also drop, and their working status may or may not reflect in new jobs numbers.

    I’m not incorporated at the moment and have never had the BLS question me regarding my self-employment status. I’m pretty sure they count me as not participating, i.e. voluntarily unemployed.

  • Icepick

    Reynolds, you are partially correct.

    Two surveys are done to create the monthly report. The establishment survey is just what it sounds like: A survey of business establishments. This survey (with one addition) provides the jobs added/lost number. This month it was 120,000. This survey must be incomplete. It cannot survey businesses that have gone out of business, nor can it survey brand new businesses. To account for this, the BLS uses something called the “birth/death model” which provides an estimate on how many jobs would have been added or lost due to the creation and elimination of businesses.

    The B/D model creates all kinds of headaches. The number that is created is blended into the number from the establishment survey, but not in a straightforward manner. Say that the B/D model creates a number of 50,000 jobs. (That is a made up number.) That doesn’t mean that 50,000 jobs get added to the BLS establishment survey data. A “black box” computer model is used that changes that number. So you can’t just take that number off the top. Calculated Risk is always (rightfully) harping on this issue, as many websites and commenters just subtract the one from the other. One even sees the same commenters doing it again and again in CR’s own comment section. Zero Hedge does this all the time too. This is all the BLS’s fault, as they should only tell us what the bottom line number is, but they don’t do that.

    Another headache the B/D model causes is that it is very inaccurate when something exceptional is happening, good or bad. So when the economy tanked in 2008 and 2009, the B/D model kept reporting happier numbers than anyone could believe. To fix this problem, the BLS revises that number four times a year based on ~something~. (They used revise the numbers only once a year, but changed in 2011.)

    The other survey is the household survey. The BLS surveys a butt-load of households, and asks them questions. From this survey we get the UE rates, the labor participation rate and whatnot.

    These two surveys can sometimes get out of step with each other, but generally they tend to trend together long term.

    So the BLS SHOULD capture some sampling of unincorporated self-employed people.

    However, there are some additional points. First, I am not entirely certain how the BLS classifies such people, so it may be that you are completely correct.

    Second, I know of some people that have taken to claiming they are unincorporated self-employed (“consulting”) so as to remove that ugly gap from their resumes.

    Third, such people might not be getting caught in the survey net. One would expect some noise on that front from month-to-month.

    Finally, I will ask again: Has anyone here ever been part of a BLS household survey? They survey a large amount of people (I’m too lazy to look up the number at the moment), and they constantly rotate new households in to and out of the survey group. But I’ve never met anyone who actually HAS been surveyed, to the best of my knowledge. (I started asking about this a couple of years ago.)

  • Icepick

    Ben, if you haven’t been surveyed you aren’t getting counted at all. They are infering everything from their surveys, so you are only a hypothetical unsampled point in the space from their perspective

Leave a Comment