About that Drop in the Unemployment Rate…

December 2nd, 2011 at 2:06 pm

As my colleague Chad points out in his analysis of today’s jobs report, the share of the working age population in the workforce fell last month, and that’s one reason the jobless rate fell from 9% in October to 8.6% last month.  See here also.

 In fact, as Chad notes, the LFPR (labor force participation rate) is down a half a percent since last November, in part because of folks giving up on the weak job market.  In that regard, the decline in unemployment is less exciting than you might think–the labor force fell by over 300K last month, accounting for half the 0.4 point decline in the jobless rate.

What if these dispirited labor-force-leavers had stayed in the job market as unemployed persons?  Then, as the figure below shows (last bar), the unemployment rate would be 9.3% instead of its current level of 8.6%.  What if only half of the leavers were unemployed (i.e., looking for work instead of out of the game altogether)?  Then the rate would still be 9%.

Source: BLS, my calculations

I’m not trying to bum anyone out here…just trying to put that sizable drop in the unemployment rate in perspective.  There’s an important substantive difference between an unemployment rate that’s falling because more people are getting jobs or one that’s going down because more people are dropping out of the labor market.

Print Friendly, PDF & Email

14 comments in reply to "About that Drop in the Unemployment Rate…"

  1. A-man says:

    If there was a net increase in jobs this month than wouldn’t the unemployment rate be 9% or lower if the Labor Force remained the same? I don’t quite get where the rate of 9.3% is coming from as that would mean that jobs were lost this month.

    • TA111 says:

      We need an increase of at least 175K per month to drop the unemployment rate because the eligible workforce grows each month. Anything less than that then the true unemployment rate will increase.
      Hope this helps.

    • Jared Bernstein says:

      It’s often the case that a net increase in jobs is accompanied by an increase in unemployment, especially if the job increase is small. EG, imagine a month where you create 10K jobs but unemployment goes up 100K.

      But all of that month-to-month change stuff is different than the 9.3%–all that says is if the labor force hadn’t declined–if all those who left it hadn’t done so, and if they were all unemployed, then the Nov rate would be 9.3.

  2. TA111 says:

    Very well done. We are in very bad shape on the employment front. The participation rate is lower than it’s been at 64%.
    And long-term unemployment is much worse. These are the important numbers. Strangely, if 2 million unemployed just give up looking for work-the Fed’s unemployment rate would drop to around 6%. How crazy is that?

  3. Seedy Guy says:

    About that drop in the labor force participation . . . it was entirely women. The number of men in the work force increased (by a scant 23,000). But nearly 340,000 women (net) left the work force. This (to me) is a much more interesting story than the drop in the UE rate . . . but it’s not a happy one.

  4. A-man says:

    Thanks Jared for your response. Here is another question I have looking at the BLS numbers.

    In October, the U/E rate was 9.0% which was determined by dividing total unemployed (13,897,000) by the total Labor Force (154,198,000).

    In November, the U/E rate dropped to 8.6% which was determined by dividing total unemployed (13,303,000) by the total Labor Force (153,883,000).

    The article mentions that the rate of 9.3% is essentially the new unemployed level (13,303,000) divided by the October (previous) Labor Force level of (154,198,000). However, when I do this calculation, I get the percentage of 8.8% and not the 9.3%.

    Could you explain the difference that I am getting? Thanks!


    • A-man says:

      Just to correct something — the last paragraph should have mentioned that I used the total of 13,618,000 and not the 13,303,000 in my calculation. So I did assume that the 315,000 loss in the Total Labor Force were unemployed.

      Here is the math for that :

      13,618,000/154,198,000 = 8.83%



  5. Mortlupo says:

    That’s the problem with the rate, just because it drops, it doesn’t mean that people went back to work. The Unemployment rate doesn’t show how many people are unemployed and looking for work just those collecting unemployment benefits.

  6. Rick Schaut says:

    The one positive here is that the household survey had a significantly higher number of new jobs than the establishment survey. A while back, the NBER published a paper showing that the ratio of the two surveys tends to be a reliable cyclical predictor. One of the reasons for this is that the household survey tends to capture activity, new business or independent contractors for example, that the establishment survey doesn’t.

  7. jo6pac says:

    You’ll see that number by election and yes they didn’t get a job, like me they just dropped off the charts. The new Amerika is about all that hope for change is to look for even less of anything in the near future when people like this making the future happen for citizens of the world.


  8. Asad Mahdi says:


    I sent a couple of posts that seem to have been deleted. I work on economic forecasts for government programs so I am really interested in the math behind the 9.3 percent. Please, if possible, provide me with the details behind that stat. Thanks so much,


  9. Asad says:


    Could you kindly respond to my question about your 9.3% rate. I believe you should change it to 8.8% unless you can prove to me where in BLS’s numbers you justify that percentage. I really respect your opinion and watch you on MSNBC religiously (even saw you talk about the U/E on Rachel on Friday) so I would appreciate a response. Thanks!


    • Jared Bernstein says:

      Sure. Just take the O.5 difference in the LFPRs and multiply that by the Nov11 population. That should get you 1.2m. Add that to the unemployed in Nov11 (13.3m) for 14.5m unemployed. Divide that by the simulated larger LF (64.5% * Nov population) of 155.1m–and that’s 9.3%.

      • Asad says:

        Jared — thanks so much for your response. I am really very appreciative of it. Just in case you happen to read these comments again, could you tell me why you use the 0.5 multiplier rather than just adding the 315k to the new U/E? That would seem to be an easy way of coming up with the total Unemployment number if all of those who dropped off (315k) were listed as unemployed.

        Thanks again!