A 'misclassification error' made the May unemployment rate look better than it is.

[Suburban Warrior]

Source: Washington Post

When the U.S. government’s official jobs report for May came out on Friday, it included a note at the bottom saying there had been a major “error” indicating that the unemployment rate likely should be higher than the widely reported 13.3 percent rate.

The special note said that if this “misclassification error” had not occurred, the “overall unemployment rate would have been about 3 percentage points higher than reported,” meaning the unemployment rate would be about 16.3 percent for May.

The Bureau of Labor Statistics, the agency that puts out the monthly jobs reports, said it was working to fix the problem.

“BLS and the Census Bureau are investigating why this misclassification error continues to occur and are taking additional steps to address the issue,” said a note at the bottom of the Bureau of Labor Statistics report.

Read more: https://www.washingtonpost.com/business/2020/06/05/may-2020-jobs-report-misclassification-error/

[Leghorn21]

1. Wat dis?? An..."error", you say?

Well, have yourself A REALLY SHITTY EVENING, BUNKERBOI

[bronxiteforever]

2. Kick and recommend for visibility

[Leghorn21]

3. So all those sweaty liars in the garden today pretty much look like idiots now, right

YOU BETCHA

[regnaD kciN]

55. No, because the narrative has been set...

People have already heard that "Trump's economy recovered much better than expected," and that's the way things are to them. They won't bother to read a lot of dry figures at the bottom of a "correction" story on page 5 of the next day's paper.

[Yavin4]

57. it's the Mueller report redux.

Get the headline out there and no one will recall the correction later. To this day, most people think that the Mueller report exonerated Trump.

[jayddrew]

4. Was this intentional to make the numbers look better thn they were?

Why this month does this occur? And how convenient for Trump to be able to tout numbers that 99% of Americans who even pay attention won't know aren't correct. This from the president who spent years claiming the numbers weren't reliable anyway.

[Mr.Bill]

10. Yes, remember during Obama's entire term,

the republicans were always saying, "Yeah, but the real unemployment rate is..."

Trump said it all through his campaign, too. Sometimes claiming it was as high as 30%.

[DallasNE]

28. An Order Had To Go Out

Otherwise the numbers would have looked much like April (fine, a little better) where 20.5 million jobs were lost but nothing like the blockbuster number reported. This was clearly done for the days photo op. The House needs testimony on why the change in characterization of these workers recently laid off and still employed rather than unemployed. The whole staff started to report these people differently so not a clerical error. Trump ordered up a photo op and he got one.

[IthinkThereforeIAM]

39. Bingo...

... you said it.

KICK

[progree]

41. Huh? The 2.5 M reported gain in jobs came from the Establishment Survey, not the Household Survey.

The Household Survey, not the Establishment Survey, is the one that had the misclassification error.

[dem4decades]

5. The error is believing anything Trump's Administration says.

[Wellstone ruled]

6. From day one of this Adminstration,

Trump stood in the Rose Garden and bragged about making the Employment Numbers always look great. They are just numbers and I will always promise you great numbers.

[Eugene]

7. Just tonight, Paul Krugman told us he trusts the professionalism of BLS staff.

Amateurishly done, but it looks like Trump Co. just took out another institution.

[DallasNE]

29. They Did Put The Footnote In The Report Saying

The impact of this mischaracterization was to understate the unemployment rate by about 3% making the magnitude absolutely huge. It doesn't look like a clerical error to me though. Somebody order the new, incorrect characterization.

[not fooled]

32. Didn't Krugman state

that the head or some Executive branch high up appointee in the BLS was from the Heriturd Fuckdation? i.e. a kochbot in there.

[NRaleighLiberal]

8. just feeding the trump bullshit machine, that little error....

what tremendous bullshit.

Everything trump touches dies.

[kysrsoze]

9. FFS? Then why don't they adjust or issue a correction

Disgusting.

[Igel]

18. Because it was in the original report. It wasn't an "error" per se.

If it's in the actual report, there's no correction needed. It's like having the NYT print a correction, "Correction: Yesterday the front page said it was Thursday, June 4, 2020. We'd like to correct that horrible error and say that it was correct, yesterday was Thursday, June 4, 2020. We apologize for the error of not being wrong, and would like to assure you that we won't do it again."

But they had to say something to acknowledge reality. They're interested in accuracy. But they have another purpose in mind.

To change the data processing method is a long process because their purpose isn't your purpose. They can't leave Friday with the orders, "Think about how to revise the process. We'll blue-sky on Monday, revise Tuesday, finalize on Wednesday and roll it out on Friday." Nope, can't do that. They're thinking long term, not "what can I use this for in politics today." They want the numbers for this month to be able to be compared with the numbers for the previous 20 Mays, and easily compared to the next few Mays' numbers. You can't just revise things on the fly because politics.

Their purpose means they have to collect the numbers in the same way and crunch them in the same way month after month. That's not negotiable, or at least isn't easily negotiable.

They revise their procedures every couple of decades. But the process is laborious because they still need to make sure that the numbers can be compared, before and after revision. That means they devise the new procedure, they bounce it around and pilot it for a few months. If it passes, then they run them in parallel for a few months, maybe a year. Then they figure out the relation of the old method to the new method, to ensure that they can be compared. Then the new method is introduced in parallel to train the people that actually need the information, then finally the new method is all there is.

We've had two, maybe three months of weirded-out data. They wouldn't be ready for the first field test yet. Much less publishing any results.

That leaves a best-guess ad-hoc "we know our data are wrong for these reasons, and we think this is the corrective." Might be. Might not be. But attributing ill will and duplicity to people just doing their jobs just shows hostility to strangers that haven't wronged you in the least, taking out your frustration on the wrong people.

[DallasNE]

30. Yes, This is How It Is Supposed To Be Done

"they have to collect the numbers in the same way and crunch them in the same way month after month."

That isn't what happened in May, however. Somebody had to order this change. We would need to see the order to determine how innocent it might have been. But I concur, it would not have been the rank and file going rogue.

[progree]

42. The same kind of thing happened in March and April.

March was "almost 1 percentage point", April was "almost 5 percentage points", and May was "about 3 percentage points" off.

The May report had this language. So did the March and April reports, except for the amount of the error:

If the workers who were recorded as employed but absent from work due to “other reasons” (over and above the number absent for other reasons in a typical May) had been classified as unemployed on temporary layoff, the overall unemployment rate would have been about 3 percentage points higher than reported (on a not seasonally adjusted basis). However, according to usual practice, the data from the household survey are accepted as recorded. To maintain data integrity, no ad hoc actions are taken to reclassify survey responses.

What "change" in May was "ordered"?

BTW, this doesn't affect the 2.5 M job gains number, that comes from a separate survey.

[DallasNE]

49. That Compounds The Problem

How can you count these as job gains if they weren't counted as job losses in prior months. This is like double counting the jobs. They were previously counted as employed and are now counted again as job gains. It is certainly misleading.

[progree]

50. They are two completely separate surveys -- the 2.5 million jobs gain comes from the Establishment

Survey which didn't have this classification error.

The count of Unemployed and Unemployment rate comes from the Household Survey, a completely different and completely separate survey with different methodologies, which has had the classification error for 3 months in a row now: March, April, and May.

[DallasNE]

11. I Have Been Trying To Call This To Attention All Day

This could have been the reason the March and April jobs report were revised downward by 642,000 meaning those reports, as grim as they were, were also rosy. In April the number of jobs initially reported as lost was 20.5 million. The actual number in May was likely around 16.5 million rather than the 2.5 million added, as reported today. This is more than a misclassification error. This was a colossal error that should cause somebody to lose their job. Indeed, was it deliberate just so Trump could gloat about it (inappropriately) today.

The House needs the head of BLS to testify under oath on how this error could have possibly happened. What changed from last month? Some kind of order went out.

[Hoyt]

12. Fortunately, even the financial websites have discussed the issue, acknowledging

actual rate is higher.

But, it is a bit encouraging that the rate was a bit better than expected. Similar results in Canada as well.

[John Fante]

13. That Trump would completely gloss over this is hardly surprising.

But what's the msm's excuse?! The headlines practically emmited an orange hue of positivity over the "surprisingly good" job numbers. It's like they can't wait to give Cheeto a win for something, anything. Even if it's not a win at all.

[DallasNE]

31. Horserace

MSM wants a horserace for the revenue that generates so they will prop Trump up for the next 5 months.

[gopiscrap]

14. trump's a fucking liar

I hope this is reported widely

[Botany]

15. An error?

I don't buy that. 3 points higher is a hell of a mistake.

[Lucky Luciano]

16. George Floyd will be unhappy about this looking down from above.

[Brainfodder]

19. WWGFD!

[Nevilledog]

37. Perfect!

[D_Master81]

26. Thought of this too

And here George Floyd was having such a great day. Hate to see it

[Miguelito Loveless]

17. Anybody know where this footnote appears?

There are dozens of tables for this report.

[mahatmakanejeeves]

51. The text of the BLS report is here:

I didn't provide links to all the tables, but you can go to the BLS and see them.

From the source. The BLS says that employment went up by 2.5 M. Not unemployment.

[IronLionZion]

20. The error seems legit, but the spin on it is pure GOP propaganda

since they know fully well that it will boost the stock market (it did) and they can discredit the BLS as a bunch of deep state socialists next.

[paleotn]

21. I call humongous BS

If you're releasing something this important, and you're getting numbers that don't make sense.....you don't release jack shit until you figure out what's driving it. The original numbers were cooked, and I bet some folks threatened to expose the games unless they came clean.

[angrychair]

22. Here is another link

Link with same info but no paywall:

https://www.stamfordadvocate.com/business/article/The-May-jobs-report-had-misclassification-error-15320999.php

[Blue Owl]

23. Has little Donny been playing with his Sharpie again?

n/t

[Cryptoad]

24. cooking them books

Bls bean counters... Joe needs to get rid of the whole crew

[mahatmakanejeeves]

52. Joe can't get rid of the whole crew, as they are by and large civil servants, which

is as it should be.

Only a handful are political appointees.

[not_the_one]

25. Wall Street.

"HIS" economy is based on Wall Street. He has to get the stocks going up again.

And Wall Street does nothing but respond in computer choreographed "knee jerks". The slightest "sign" is interpreted in a way that causes the knee to jerk, making more trades, earning brokers more fees.

Wall Street goes up due to the "error", so they report the "error", and it may go down some, but not as much as it went up, so they made money off of the "error".

Wall Street should be required to set up an account for every single American, so we ALL could receive the benefits of the rigged, computerized trading bonanza. We ALL should be able to sit back and make money off or our money. You never know, we may never have to work another day in our lives. After all, WE buy the products that corporations make a tonne of money off of, money coming directly from our pockets, THEN Wall Street also gets more profits from the selling of the stocks of those companies that WE make successful.

They make trillions. It would be nice if each of us got a few thousand a year off of OUR money in Wall Street. Every little bit helps.

I honestly don't know whether to include the smilie, or not...

[SergeStorms]

27. And Wall Street went nuts over the news today...

will we have an equally shitty day Monday? I wouldn't count on it. Wall Street has no basis in reality anymore. Investors seem to think there is limitless perpetual growth in the market. Sheep. More sheep (like evangelicals) that follow republicans around like they're the Pied Piper of Hamelin. One day the music is going to stop, and there's going to be a shitload of dunderheads left without chairs.

[Wuddles440]

53. Powell and the Fed.....

have flooded the market with liquidity and have instituted policies for a moral hazard. Investors are confident that they'll be bailed out by the Fed and possibly Congress no matter how reckless they get.

[SergeStorms]

56. Sooner or later...

someone is going to have to pay the piper. If they're going to just keep printing money inflation is going to catch up to us sooner or later.

[Midnight Writer]

33. Bartcop's Law: When someone makes an "error" that profits them, expect to see that same error again

and again and again.

[soldierant]

34. I understand what a "classification error" would be, but

what's a "misclassification error"? You got caught?

[Jim__]

35. Yes and the error was found just in time for the Friday night news dump.

I hope the media blasts away at this for the entire weekend and into next week. And loudly recalls this error whenever the Trumpists announce unexpected good news.

[Beartracks]

36. And 2 + 2 5.

"It's really X, but we OFFICIALLY stated that it's Y, so... It's Y."

==============

[cstanleytech]

38. Error my ass!!! More like Trump and his people trying to cook the books is my bet.

[progree]

40. March and April also had the same problem. March was "almost 1 percentage point",

April was "almost 5 percentage points", and May was "about 3 percentage points" off (i.e. the unemployment rates in the respective months would be higher by these amounts were it not for these mis-classification errors).

Using corrected values for both April and May, the unemployment rate dropped from almost 19.7% in April to about 16.3% in May.

[progree]

43. No need to hunt around for non-paywalled articles. It's right in the BLS report summary

and main report. In a big box. https://www.bls.gov/news.release/empsit.nr0.htm

(this in response to a couple of replies in this thread, not to the OP).

(I bolded the last paragraph)
_______________________________________________________________________________________
| Coronavirus (COVID-19) Impact on May 2020 Establishment and Household Survey Data
|
|
| Data collection for both surveys was affected by the coronavirus (COVID-19) pandemic.
| In the establishment survey, approximately one-fifth of the data is collected at four
| regional data collection centers. Although these centers were closed, about three-
| quarters of the interviewers at these centers worked remotely to collect data by
| telephone. Additionally, BLS encouraged businesses to report electronically. The
| collection rate for the establishment survey in May was 69 percent, slightly lower
| than collection rates prior to the pandemic. The household survey is generally
| collected through in-person and telephone interviews, but personal interviews were
| not conducted for the safety of interviewers and respondents. The household survey
| response rate, at 67 percent, was about 15 percentage points lower than in months
| prior to the pandemic.
|
| In the establishment survey, workers who are paid by their employer for all or any
| part of the pay period including the 12th of the month are counted as employed, even
| if they were not actually at their jobs. Workers who are temporarily or permanently
| absent from their jobs and are not being paid are not counted as employed, even if
| they are continuing to receive benefits.
|
| The estimation methods used in the establishment survey were the same for May as they
| were for April. However, after further research, BLS extended the modifications that
| were made to the April birth-death model back to March, which accounted for a portion
| of the revision to March data. For more information, see
| www.bls.gov/cps/employment-situation-covid19-faq-may-2020.pdf.
|
| In the household survey, individuals are classified as employed, unemployed, or not
| in the labor force based on their answers to a series of questions about their
| activities during the survey reference week (May 10th through May 16th). Workers who
| indicate they were not working during the entire survey reference week and expect to
| be recalled to their jobs should be classified as unemployed on temporary layoff. In
| May, a large number of persons were classified as unemployed on temporary layoff.
|
| However, there was also a large number of workers who were classified as employed but
| absent from work. As was the case in March and April, household survey interviewers
| were instructed to classify employed persons absent from work due to coronavirus-
| related business closures as unemployed on temporary layoff. However, it is apparent
| that not all such workers were so classified. BLS and the Census Bureau are
| investigating why this misclassification error continues to occur and are taking
| additional steps to address the issue.
|
| If the workers who were recorded as employed but absent from work due to "other
| reasons" (over and above the number absent for other reasons in a typical May) had
| been classified as unemployed on temporary layoff, the overall unemployment rate
| would have been about 3 percentage points higher than reported (on a not seasonally
| adjusted basis). However, according to usual practice, the data from the household
| survey are accepted as recorded. To maintain data integrity, no ad hoc actions are
| taken to reclassify survey responses.
|
| More information is available at
| www.bls.gov/cps/employment-situation-covid19-faq-may-2020.pdf.
______________________________________________________________________________________

[maddogesq]

44. Hate to keep harping on this...

but look at the U6. It’s down about a point to 21.2%. This one is the real deal and a more accurate reflection of what’s going on.

https://www.bls.gov/news.release/empsit.t15.htm

[progree]

45. And its also low by about 3 percentage points, since the U3 unemployed is a subset of U6's unemployd

Likely more, since almost certainly some of those in U6 that aren't in U3 were also misclassified. OTOH, the denominator of U6 is larger than U3 (its complicated). But anyway, add 3% to 21.2% and we're at 24.2%.

[Hugin]

46. I personally know two people who lost their jobs in the period.

Up from zero... Ever.

I don't trust any numbers on anything coming out of the Feds right now and even less the numbers from Wall Street.

[Karma13612]

47. I suspected that all along.

The fudged numbers caused a spike in the stock market. Since tRump relies on the market to fuel his reelection support, I am convinced the jobs report was fiddled with.

Surprise surprise.

[JI7]

48. 13.3 isn't anything to celebrate

[regnaD kciN]

54. "Misclassification Error"...? Sounds like the political "Wardrobe Malfunction"...

[Yo_Mama_Been_Loggin]

58. So this is what Donny Dotard is crowing about?

Lowered expectations I guess.

[BadGimp]

59. That liddle "misclassification" caused the Unemployment Rate to jump 24.8%

How perfect the timing right? Likely an honest mistake on the part of a lot of people right?

WTF did they hide to cause this?

[ck4829]

60. Kick

[ck4829]

61. Kick

[area51]

62. "... this misclassification error continues to occur...."