Friday, August 7, 2009

Unemployment Rate Drop = Decline in Labor Force

The country shed another ~250k jobs in July, but the unemployment rate dropped to 9.4%. How is that? It looks like individuals have left the labor force.

Month over Month Change (S.A.)

How does this work? The numerator in the unemployment rate is unemployed... the denominator is labor force. Simplified example:

  • 19 people are unemployed out of 200 in the labor force = 9.5% unemployment rate
  • 1 of those unemployed individuals leaves the labor force
  • 18 people are unemployed out of 199 in the labor force = 9.0% unemployment rate
This also explains why the percent of those unemployed decreased at a greater rate than the labor force (1/19 = a 5.2% drop in the # unemployed, but 1/200 = a 0.50% drop in the labor force).

So, was the data better than expected? Absolutely (good to great news on a relative basis), but lets make sure we understand the math... and the continued trend.

Employment as a Percent of Entire Population

Source: BLS


  1. ["So, was the data better than expected? Absolutely (good to great news on a relative basis), but lets make sure we understand the math..."]

    Expected by who ??

    Why are the "expectations" of that 'who' ... of any practical value at all ?

    What is the track-record of accuracy for that 'who' ?

  2. i'll throw a "me" out there as to whom it was better than expected, but i understand your point (and i have no accuracy). if it makes it more palatable, change the expected to last month.

    ["So, was the data better than last month? Absolutely (good to great news on a relative basis), but lets make sure we understand the math..."]

  3. I'm not one to blame illegal aliens for 95% of the things they're blamed for, so don't get me wrong. However, the headlines have been consistent that illegal immigrants have been returning to their home countries as a result of the poor employment situation here in the US.

    This raises an important question: Who is leaving the employment pool? Does this meant that unemployment numbers have been inflated by the presence of illegal aliens?

  4. i don't believe illegal immigrants are in the survey. my reasoning is that the # of those labeled 'not in labor force' has increased. if they fled, they wouldn't be in the figures at all...

  5. Anon - it could be that LEGAL immigrants have been laid off and/or cannot find employment in the US and are leaving. That, I would think, would be in the survey.

  6. Three thoughts for the BLS-derived graph of "Employment as a Percent of Entire Population"
    - 5+ yrs to work our way through the significant troughs (Aug 89-Aug94 is a good example)
    - The rise to near 65% employment can be coorelated, at least loosely, to the dot-com rise (and subsequent fall). That begs the question: what new industry will pull us out of this slump?
    - Finally, I recall in an MBA course to beware of government changes to the rules by which each data point is calculated. Here, for example, the BLS site notes "Updated population controls are introduced annually with the release of January data" What are the updated population controls this year and more importantly what else has changed that we don't know about?

  7. Jake - why did you choose to leave out the 1970s and early 80s which would have likely shown the recent #s aren't out of line?

  8. i was just trying to show that we are at our 25 year low to put this into perspective. didn't mean to cut off any data intentionally, but i can see where it would seem that way, so i do apologize.

    that said the 1960's - 1970's data is too stale. that was when the two-income family didn't truly exist (i.e. when women weren't necessarily expected to work and/or given the same opportunity to do so). thus not remotely an apples for apples comparison.

  9. this also explains why weekly hours worked were 5 more hours per week than current levels throughout the 1960's...