Showing posts with label Establishment Survey. Show all posts
Showing posts with label Establishment Survey. Show all posts

Saturday, September 2, 2023

The unemployment rate rose to 3.8%, but not because people lost jobs

 The unemployment rate rose to 3.786% from 3.495% on a bigger 736k increase to the size of the labor force than to the employment level, not because people lost jobs.

The employment level actually made a new high in August 2023, but up a smaller 222k. 

The unemployment rate went up in August because record new high employment in August, 161.484m, is a smaller percentage of a new larger labor force in August, 167.839m than was the case in July: 96.2% in August vs. 96.5% in July = 3.8% and 3.5% unemployed respectively. 

And do not mix the limited Establishment Survey (122,000 businesses and agencies) total nonfarm jobs oranges (156.419m) with the unemployment rate Household Survey (60,000 households) whole universe of jobs apples and try to make them agree. They don't, and never will.

The Establishment Survey went up 187k in August, but the unemployment rate is not derived from that survey. 

 



 

Saturday, January 18, 2020

CNN Business is stupid: "Women now hold more jobs than men"

Here's the lede:

Women held slightly more jobs than men in December — the first time that's happened in nearly a decade. 

That's based on the Establishment Survey, which undercounts civilian employment by the millions.

The Household Survey shows 84 million men employed in December 2019, 74.8 million women. Employed women have never outnumbered employed men in the entire post-war history of the data.

Thursday, August 22, 2019

Meanwhile benchmark revisions to the quick and dirty monthly Establishment Survey indicate 501,000 jobs have to be SUBTRACTED through March 2019

That is expected to bring down 2018 job creation to roughly 181,000 monthly instead of 223,000.

I say, so what? 223,000 monthly wasn't indicative of a jobs boom anyway.

Jobs numbers under Trump are like Trump Steaks, overcooked.

Saturday, July 6, 2019

Headline payrolls in 2019 may be overstating the real numbers by more than 25%



Months from now, the Establishment Survey will undergo its annual retrospective benchmark revision, based almost entirely on the Quarterly Census of Employment and Wages conducted by the Labor Department. ...

The latest QCEW data are available through 2018, but note how much worse the 2018 QCEW data look than the Establishment Survey data, even though the two appear fairly similar in previous years, for which the latter has already undergone the requisite revisions. The Establishment Survey’s nonfarm jobs figures will clearly be revised down as the QCEW data show job growth averaging only 177,000 a month in 2018. That means the Establishment Survey may be overstating the real numbers by more than 25%.

Wednesday, February 6, 2019

Kellyanne Conway baldly lies on The Rush Limbaugh Show, claims over 7 million jobs created

You don't lose control of the US House after creating over 7 million jobs in just over two years.

These people are blind, but the voters are not.

There is NO measure which comes even close to over 7 million new jobs.

Total nonfarm payrolls (Establishment Survey) between Nov 2016 and Jan 2019 are up 5.3 million, seasonally adjusted, 1.7 million not seasonally adjusted.

Civilian employment (Household Survey) is up 4.5 million seasonally adjusted, 2.6 million not seasonally adjusted.

The sum of usually full time and usually part time (Household Survey) is up net 4.5 million seasonally adjusted, 2.6 million not seasonally adjusted (mirrors civilian employment level).

The employment-population ratio is up from 59.8% to 60.7%, an increase of 4.6 million jobs from 152.2 million to 156.8 million.

And 78% of these new jobs have gone to minorities, not to the white majority.




CONWAY: . . . He’s going to talk about the booming economy, and it is the Trump economy. Over seven million jobs created, wages are up, unemployment is down, small business formation is going well. People are just… They’re back to work. We’re a country that works.

Wednesday, January 31, 2018

Sorry, Trump didn't create 2.4 million jobs "since his election", and even if he did, that sucks

Not even the fact-checkers seem to want to get this right. Is the whole country taking stupid pills?

Total nonfarm, not seasonally adjusted, stood at 146.393 million in November 2016. In December 2017 the level was 148.346 million, an increase of 1.95 million.

Seasonally adjusted, the increase went from 145.170 million to 147.380 million, an increase of 2.21 million.

The former figure is 150,000 per month (13 months), the latter 170,000 per month. Meanwhile Trump is claiming 184,462 per month.

These are terrible numbers, including Trump's, which however also appear to be cooked on a bonfire.

In a booming economy, monthly increases well above 200,000 are indicated. That's what we got under Reagan and Clinton, but not now, not by a long shot.

Turn away from these Establishment Survey numbers and consider the Household Survey figures and the picture looks even worse.

Not seasonally adjusted the sum of usually full-time and usually part-time is up just 1.216 million in 13 months. This has been seasonally adjusted up to barely 1.9 million. We're talking 93,500 per month to 146,000 per month, quite the spread. Not exactly confidence inspiring numbers.

The truth is that employment gains have gone soft in 2017 compared with 2016, down about 15%. In the 13 months up to November 2016, total nonfarm jobs seasonally adjusted increased at a monthly pace of not quite 200,000 which was nearly 18% better than under Trump so far. 

Trump better hope hiring picks up soon, or this charade will quickly be seen for what it is, all hat and no saddle.

Still waiting for the boom.


Friday, April 7, 2017

ADP had private payrolls up 263,000 earlier this week, but we only get 98,000 from the BLS today

When the expected BLS increase to total nonfarm was 180,000.

What's up with that?!

Long story short: Forget ADP, and employment gains have slowed by 15% in the last year. 

ADP is designed to try to predict what the BLS is going to say, and is known to fail at this. It is questioned why ADP even bothers. I agree.

A government measure from the Household Survey, the sum of usually full-time and usually part-time, not seasonally adjusted, is up only 1.89 million (157,500 per month) year over year in March 2017.

Total nonfarm, on the other hand, from the Establishment Survey, is up 2.135 million, not seasonally adjusted (178,000 per month) year over year.

That last number, 178,000 per month, is what the BLS also says is the current 3-month average in March 2017.

Compare that to March 2016 when the 3-month average was 209,000.

There's the truth.

The soft number of 98,000 in the headline today contributed to the climb-down in the 3-month average this month. It will be revised next month, probably up because it's so low. But it's the revisions down for the previous two months totaling 38,000 which are the clue to look farther back.   

The absolute number is not important because we can't be certain about it. It is only an estimate anyway, not an actual count. But we can be certain that the long-term behavior of the estimates shows an employment slow down of about 15% year over year.

It was well underway before Trump even got elected.

You can see that in full-time employment gains. Measured in March year over year, full-time gains peaked in March 2015 at almost 3 million additions. Gains have steadily fallen since, to 2.5 million in March 2016 and just 2 million now in March 2017.

And the bottom line there is we've added about only 4.7 million full-time jobs since March 2008 on net.

After NINE years.

The country remains mired in shrunken conditions from which it has not escaped.

In a real recovery, the 10 million full-time jobs lost in 2009 and 2010 would have been fully replaced by 2013 at the latest. It took until 2015, and the momentum immediately started to recede.

Friday, December 5, 2014

Usually part-time vaults 465,000 in one month to new all-time high of 28.225 million

Those who work usually part-time vaulted 465,000 month over month, not-seasonally-adjusted, in today's Household Survey data in the Employment Situation Summary for November 2014. That puts the metric at an all-time high of 28,225,000, about 100,000 higher than the previous peak reached in the wake of the late economic depression.

Meanwhile, those who work usually full-time dropped 735,000 from October to November.

Full-time usually peaks in the summer and part-time usually peaks in the winter, so the data coheres with past experience, except the new high in part-time is a little worrisome.

Voluntary part-time is up big month over month (536,000) while involuntary part-time is down (74,000). Is that a sign of acquiescence to a new normal of part-time work? Admittedly, involuntary part-time is still 2.5 million higher than it was in the autumn of 2007, but it has fallen 2 million between 2011 and 2014 even as today there are 2.5 million more full-time jobs than there were a year ago.

Full-time remains 3.8 million under the 2007 peak.

Meanwhile multiple job holding is down 224,000 month over month, and the total employed is actually down 270,000, as is the total number unemployed, down 50,000 not-seasonally-adjusted.

It looks as if the big jump of 321,000 in total nonfarm from the Establishment Survey is a phenomenon of part-time. Whether these part-time jobs become an enduring phenomenon in the form of permanent jobs won't be clear until after the new year.

The unemployment rate at 5.8% remains where it is as those not in the labor force continues ever upward, this time 536,000 higher from October to November. The metric hovers near the all-time high of 92.5 million reached in April. People dropping out means fewer people to count as unemployed.

The civilian labor force shrank in size 319,000 from October to November.

Friday, November 7, 2014

Unemployment falls to 5.8%, 214,000 jobs added in October

Average jobs added monthly in the last twelve months rose to 222,000. A year ago at this time 190,000 were being added monthly in the prior twelve months. During the Reagan boom 250,000 were added monthly for six years. During the Clinton boom 235,000 were added monthly for eight years. The 17% increase in the pace in the last year is a good thing, but we've got a long way to go, if it can even be sustained. A different indicator may give reason to hope so.

From 2008 to 2013, the percentage of the work force participating had been in steady decline measured October to October, until today. The labor participation rate now is 63.0% vs. 62.9% a year ago, not seasonally adjusted. That's not much but it may mark a turning point. It remains to be seen if the 62.5% level reached in January was in fact the bottom.

Looking at the broad measures, those who say they work usually part-time are up 414,000 not seasonally adjusted from a year ago, but the level remains 233,000 off the previous peak for an October, which occurred in 2012.

Those who work usually full-time are up an astounding 3,378,000 not seasonally adjusted from a year ago at this time. Compared with the peak year of 2007 for this metric, October on October, those who work usually full-time today are still 1.83 million fewer in number than then, not seasonally adjusted.

If you add the two categories together and divide by twelve, you get 316,000 jobs added monthly, not 222,000 as stated in the Establishment Survey which takes a larger sampling.

Go figure.

Average hours went up .1 and average earnings went up 3 cents.