3 Misconceptions About the Jobs Report

Party time!

Tomorrow morning, the Bureau of Labor Statistics will report May's monthly jobs report. This is a widely followed, highly anticipated report that can move the Dow Jones (INDEX: ^DJI  ) in a big way one direction or the other.

It's also one of the most abused and misinterpreted reports around. Here are three of the biggest whoppers I run into frequently.

Misconception No. 1: "Jobs numbers missed expectations."
March's jobs report showed the economy created 120,000 jobs. The average forecast was for 205,000 jobs. Not good, right? "Stocks decline as jobs report misses estimates," reported Bloomberg.

But statistically, that's not quite right. Every monthly jobs report has a margin of error of plus or minus 100,000, so there's no meaningful difference between 120,000 and 205,000.

Why so much wiggle room? The surveys used to calculate jobs reports don't actually count every person in America -- that would be too hard for anyone to do every month. Instead, they use small samples, and extrapolate the results of those samples out to reflect how many jobs were probably created and what the unemployment rate probably is. Since there's a risk the bureau might survey a group of people not reflective of the entire nation, that extrapolation can't be done with precise numbers; just broad ranges.

The BLS gives an example. When it reports that 50,000 jobs were created in a month, what it's really saying is that there's a 90% chance that the monthly change was between -50,000 and +150,000.

Likewise, the margin of error for the unemployment rate is 0.2%. So when it's reported that the unemployment rate is 8.1%, that basically means there's a 90% chance that it's between 7.9% and 8.3%.

Nearly everyone looks past this, so if the jobs report misses estimates by even a few thousand, it gets called as a "disastrous miss" by the media, though it might be nothing of the sort. We really don't know if 120,000 jobs were created in March -- just that we're reasonably confident it was between 20,000 and 220,000.

Most initial jobs reports are also revised in subsequent months, sometimes substantially. The revisions go both ways: In late 2009 we learned that job losses from early in the recession were worse than originally reported, and over the last nine months we've learned that recent jobs gains have been stronger than originally reported.

The takeaway? Find some salt, and take a big, healthy grain of it.

Misconception No. 2: The government's unemployment rate doesn't include those who have given up and stopped looking for work.
While testifying before Congress last year, a senator complained to Federal Reserve chairman Ben Bernanke that the real unemployment rate was higher than it looked. Alternative measures that include those who have stopped looking for work, the senator noted, show the unemployment rate is much higher than what's reported by the BLS.

Bernanke politely pointed out that, actually, those alternative measures are reported by the BLS.

The government calculates unemployment six different ways. The one most often used in the media is called U3, and is calculated by taking the number of people without jobs who have actively looked for work in the last four weeks as a percentage of the working-age population. That rate is currently 8.1%.

Another version of the unemployment rate called U6 includes a bunch of other people -- those who have become discouraged and stopped looking for work, and those who are working part-time but want full-time work. That rate is now 14.5%.

Both figures are reported clear as day in the monthly unemployment reports. It's up to the analyst or journalist to decide which to use.

And, importantly, every version of the BLS' unemployment rate has declined over the last few years:

Source: BLS.

Misconception No. 3: Those who aren't eligible for unemployment benefits aren't counted in the unemployment rate.
This is a frequent criticism, and it's just plain wrong. Being counted in the unemployment rate has nothing to do with whether or not you're receiving unemployment benefits.

The unemployment rate (or rates) is calculated by a monthly survey of 60,000 households, covering 110,000 people and more than 2,000 separate regional areas.

The survey asks dozens of questions, but the two most important for determining the unemployment rate are: Did you work for pay or profit last week? If no, have you been doing anything to find work during the last four weeks? If yes, you're unemployed. End of story. It has nothing to do with unemployment benefits.

The government does report on the number of Americans receiving unemployment benefits, but it's totally separate from the surveys tracking the unemployment rate.

Now get ready for the excitement!

Fool contributor Morgan Housel doesn't own shares in any of the companies mentioned in this article. Follow him on Twitter @TMFHousel. The Motley Fool has a disclosure policy. We Fools may not all hold the same opinions, but we all believe that considering a diverse range of insights makes us better investors. Try any of our Foolish newsletter services free for 30 days.


Read/Post Comments (26) | Recommend This Article (28)

Comments from our Foolish Readers

Help us keep this a respectfully Foolish area! This is a place for our readers to discuss, debate, and learn more about the Foolish investing topic you read about above. Help us keep it clean and safe. If you believe a comment is abusive or otherwise violates our Fool's Rules, please report it via the Report this Comment Report this Comment icon found on every comment.

  • Report this Comment On May 31, 2012, at 6:10 PM, xetn wrote:

    Great spin Morgan. Do you work for the BLS?

    Now for the real stats:

    http://www.shadowstats.com/alternate_data/unemployment-chart...

  • Report this Comment On May 31, 2012, at 6:11 PM, TMFMorgan wrote:

    That data includes people who haven't look for a job in a year or more. Frankly, if you haven't even looked for a job in a year, you don't want on. You're not unemployed.

  • Report this Comment On May 31, 2012, at 6:35 PM, Wade32ru wrote:

    Great review on sampling statistics. Thanks!

  • Report this Comment On May 31, 2012, at 7:27 PM, StillWaters43 wrote:

    Those survey questions you reference are the same questions that are asked to determine if you get future bi-weekly unemployment checks. So, if this is the survey source it only goes to those collecting benifits?.

  • Report this Comment On May 31, 2012, at 7:32 PM, TMFMorgan wrote:

    ^ Unemployment benefits are administered through the states. BLS is federal. So, no.

  • Report this Comment On May 31, 2012, at 9:58 PM, SissyTheClown wrote:

    Have you ever thought, TMFMorgan, the reason those people haven't looked for a job in a year or more is because they are all to aware that there are no good-paying jobs left for middle class workers? They've all been shipped out of the country because of the cheaper labor and much lower tax rates. Our government has shot us all in the foot with these outrageous taxes on businesses.

  • Report this Comment On May 31, 2012, at 10:13 PM, TMFMorgan wrote:

    <<Have you ever thought, TMFMorgan, the reason those people haven't looked for a job in a year or more is because they are all to aware that there are no good-paying jobs left for middle class workers?>>

    How can you be all too aware if you're not even looking? And you're not unemployed if you're only looking for "good-paying jobs," which implies there are other jobs available, but not up your standards. Anyone at any time can be "unemployed" if they set their standards high enough and then don't even bother trying.

    <<Our government has shot us all in the foot with these outrageous taxes on businesses.>>

    Corporate tax revenue was $370 billion in 2007 (when unemployment was 5%), and $181 billion in 2011 (when unemployment was 9%). Corporate profits grew during that period, so it wasn't from a lack of activity.

    Corporate taxes as a % of GDP have dropped like a rock over the last 50 years:

    http://research.stlouisfed.org/fred2/graph/fredgraph.png?&am...

  • Report this Comment On May 31, 2012, at 11:16 PM, JULPAC wrote:

    How is this "party time?"

    While I’m supposed to receive this article in a way as to not mismanage my accounts based on the jobs report set to be out tomorrow I can’t help but notice the slight political slant in favor of this administration w/in it. This quiet preference is further observed in your responses to other readers comments

    How is it possible for anyone to get "excited" when you deem them as "not unemployed" simply because they refuse to take any job that is availalble? Of course there are plenty of jobs for us Americans to take up - we could easliy apply @ the nearest fast food chain, or if we're willing to get down & dirty we could take up the positions that migrant workers from down South travel hundereds of miles for just to pick up our crops!

    Yes your completely right - they're "not unemployed, they just don't want on.(!)"

  • Report this Comment On May 31, 2012, at 11:45 PM, SuntanIronMan wrote:

    "It's up to the analyst or journalist to decide which to use."

    Or politician, deciding which to spin to the American people. The numbers that most favorable to your party or the numbers most unfavorable to the other party (and if you are a 'good' politician, you'll spin the numbers to accomplish both at the same time).

  • Report this Comment On June 01, 2012, at 3:10 AM, Analog80 wrote:

    There's one other significant number that's not being included.

    The people in prison. The last time I checked if you included the people in prison it would add at least a full 1% percent+ to the unemployment number.

  • Report this Comment On June 01, 2012, at 9:30 AM, dragonmonkey wrote:

    Good article. Stats are stats...removing the b.s. and laying them out for what they are makes them more usable and valuable. Not sure why people choose to criticize you for explaining what the numbers mean. I don't see a political slant to your article. I also don't understand the critics here equate someone who has made no attempt to look for a job whatsoever in over a year with the strawman argument of migrant workers picking crops. Get real. U6 is about as fair a number as you can expect, and the other numbers have their uses too.

  • Report this Comment On June 01, 2012, at 9:50 AM, TMFCop wrote:

    Morgan,

    The one quibble I'd have with your article is your your statement that BLS lays out the real numbers "clear as day." In fact, the actual BLS press release highlights only the U3 numbers; you have to go to the appendices to find those numbers.

    As your link proves, the U6 numbers are contained in Table A-15, which is No. 5 on a list that includes 29 other appendices. Although the media can be witless, you can't really blame them for highlighting the number BLS shouts from the rooftops.

    Moreover, it's also clear BLS doesn't even believe in the numbers it reports. If you go to yet another appendix, the Technical Note, you'll see that despite purportedly have high 90% confidence in its numbers, they can be off by as much as 100,000 jobs on average. That means the 69,000 jobs created in BLS's current estimate -- and it is an estimate as there are all sorts of gymnastics that go into massaging these numbers -- it could really mean it was a loss of as much as 31,000 jobs in May.

    So again, BLS issues a release that touts the U3 numbers, and while it does offer up the U6 numbers too, they're buried in footnotes and appendices which hardly makes it as "clear as day " to find.

    Rich

  • Report this Comment On June 01, 2012, at 9:58 AM, TMFMorgan wrote:

    First-world problems: clicking on a link once a month is too hard for journalists to do. You could just as easily complain that the monthly report itself is buried on an obscure government website.

    <<Moreover, it's also clear BLS doesn't even believe in the numbers it reports.>>

    It's not that it doesn't believe them; this is just how statistics of samples works. BLS didn't invent extrapolation statistics.

    <<it could really mean it was a loss of as much as 31,000 jobs in May.>>

    Or a gain of 169,000 ... right?

  • Report this Comment On June 01, 2012, at 11:13 AM, Turfscape wrote:

    Morgan, once again you've made the fatal error of not politicizing something. This is America, gosh darn it all...you should take a topic like unemployment statistics and use it to further the agenda of a single political party. The fact that you seek greater understanding of a situation and try to relate it to investing proves that you are, in fact, a freedom-hating, pinko commie, socialist.

  • Report this Comment On June 01, 2012, at 11:22 AM, TMFCop wrote:

    <<<<it could really mean it was a loss of as much as 31,000 jobs in May.>>

    Or a gain of 169,000 ... right?>>

    Which really says something about the value of their numbers. I'm wondering why we even pay attention to those numbers when you can drive a tractor trailer through an opening that large.

    But it's not just about extrapolation statistics. The numbers are based on surveys mailed out to sample employers, which BLS says is about 140,000.

    What it <i>doesn't</i> say is the percentage of surveys that are completed and returned. We really don't know what the sample size is that BLS is basing its numbers on and then massaging on oftentimes bogus seasonality, etc. (yeah, when Census Bureau workers are included that should be accounted for; a lot of its other "seasonality" changes aren't anything other than nonsense).

    So BLS isn't being exactly transparent, particularly when there IS available real-time data in the withheld income & employment taxes sent to Treasury everyday by all employers. So there isn't any need to extrapolate any data at all: they can get real numbers right away if they cared to look.

    Rich

  • Report this Comment On June 01, 2012, at 11:27 AM, StopPrintinMoney wrote:

    Are you trying to say things are much better than reported?

  • Report this Comment On June 01, 2012, at 11:39 AM, TMFMorgan wrote:

    <<Which really says something about the value of their numbers.>>

    Again, BLS didn't invent confidence intervals. Open any high-school stats textbook. This is standard stuff.

    <<a lot of its other "seasonality" changes aren't anything other than nonsense).>>

    Then look at nonseasonally adjusted figures. Those are reported. And seasonality doesn't matter over the course of a year ... if jobs are overreproted in one month, they're underreported in other months.

    <<So BLS isn't being exactly transparent, particularly when there IS available real-time data in the withheld income & employment taxes>>

    There's a reason payroll tax receipts are bad at measuring jobs growth. Let's say you get your pay cut from $60k to $50k. Suddenly, you're paying less in payroll taxes. But are you unemployed? no. Same if you get a raise from $50k to $60. Now you're paying more in payroll taxes. But was another job created? No. Also, the recent payroll tax cut makes payroll tax receipts look far worse now compared with years prior.

  • Report this Comment On June 01, 2012, at 11:46 AM, TMFMorgan wrote:

    <<Are you trying to say things are much better than reported?>>

    No.

  • Report this Comment On June 01, 2012, at 1:51 PM, astuber9 wrote:

    However you slice these numbers, they are bad. It scares me that the government is running big deficits to try to spur growth and this is what we have. Can you imagine what kind of recessions we will have when the government starts paying off debt?

  • Report this Comment On June 01, 2012, at 1:56 PM, TMFMorgan wrote:

    ^ For what it's worth, government spending is falling for the first time in 60 years:

    http://research.stlouisfed.org/fred2/graph/fredgraph.png?&am...

  • Report this Comment On June 01, 2012, at 3:28 PM, astuber9 wrote:

    I will take any spending decrease, but you are kind of cherry picking with comparing the last couple of years with the binge spending of 09-10 from bailouts, really high unemployment, etc.

  • Report this Comment On June 02, 2012, at 8:14 AM, svcbearcat72 wrote:

    <Then look at nonseasonally adjusted figures. Those are reported. And seasonality doesn't matter over the course of a year ... if jobs are overreproted in one month, they're underreported in other months.>

    A point of information - it is "seasonally adjusted" figures not "nonseasonally adjusted". If there was no seasonality, there would be no need to adjust the data.

    Additionally, seasonality does matter over the course of the year. Its magnitude and influence on the unemployment rate is easily seen and its effects are easily measured.

    Also, your comment about over-reporting and under-reporting has nothing to with seasonality or seasonal adjustment. That is simply tied to the BLS early reporting and subsequent revisons that follow

  • Report this Comment On June 02, 2012, at 12:00 PM, ShaunConnell wrote:

    " there's no meaningful difference between 120,000 and 205,000."

    Actually, yes there is. It's an average or a middle ground, and you look for the trends. The trend, in light of the rest of the world slowing down economically, isn't good.

    People who use it as the gospel are goofy, but let's not kid ourselves and say it's not "meaningful" just because the political implications might not be what we were after.

  • Report this Comment On June 06, 2012, at 10:21 AM, Darwood11 wrote:

    If I look at the "trend" then I must conclude that it will require about a decade of steady improvement of the type we have been experiencing to get "joblessness" back to what we would have considered "normal."

    Since we're about 4 years into this jobless cycle that implies about 6 more years if steady economic improvement should occur.

    Of course there is the possibility of a number of disruptors which could exacerbate or improve the situation.

    However, as recessions seem to occur with regular frequency it would be my opinion that it is more likely that unemployment will "flatten" for a time before continuing downward.

    As for the current numbers, we all know that the BLS issues updates and corrections and so I am always skeptical about the most recent numbers.

  • Report this Comment On June 06, 2012, at 10:25 AM, Darwood11 wrote:

    If I look at "trends" then I suppose I should also consider the winding down of the wars overseas and probable release of military personnel, the continued lower spending rate at local governments to coincide with tax collections, and so on.

    In other words, it would seem that jobs in the government sector will continue to be pared.

    Finally, it is possible that our Congress won't come to a compromise about the deficit and that certain automatic cuts will occur in the military. That will release a lot of workers in the industries that support the military and aerospace sectors.

    As Morgan said "Now get ready for the excitement!" Yes, indeed.

  • Report this Comment On June 14, 2012, at 1:53 AM, thidmark wrote:

    When the jobs report is terrible and two previous months are revised down, you have a pretty good idea of what's going on.

Add your comment.

Sponsored Links

Leaked: Apple's Next Smart Device
(Warning, it may shock you)
The secret is out... experts are predicting 458 million of these types of devices will be sold per year. 1 hyper-growth company stands to rake in maximum profit - and it's NOT Apple. Show me Apple's new smart gizmo!

DocumentId: 1900004, ~/Articles/ArticleHandler.aspx, 12/21/2014 9:16:18 PM

Report This Comment

Use this area to report a comment that you believe is in violation of the community guidelines. Our team will review the entry and take any appropriate action.

Sending report...


Advertisement