Quote:
Originally Posted by middle-aged mom
Project 538 does a decent job of explaining employment numbers.
The number that gets the most attention is an ESTIMATE derived from surveying 150,000 employers and government agencies.
|
Actually, no, Project 538 is wrong. The number that gets the most attention is the fake and phony seasonally adjusted data from the Current Population Survey, not the Current Employment Survey.
The data from the Current Population Survey is data manipulated by the X-13ARIMA-SEATS software program that you can download for free at the US Census Bureau website.
To use the program, you enter the number from your data then add weighting factors. You can add weighting factors for anything you want.
So, let's assume 1,537 people died in traffic accidents in July. The actual number may need to be revised later because some people were only injured but they will die of their injuries in the weeks and months following.
We enter 1,837 into the X-13ARIMA-SEATS software program then add our weighting factors, like temperature, weather conditions, holidays, you saw two birds in a tree or whatever you want.
Click on "Run" an
viola! We get 1,321 Seasonally Adjusted Traffic Deaths.
What happened to the other 616 people who died of traffic deaths in July?
Who cares? Besides, they're dead so it's not like they're going to whine about it.
People like the OP and the Media will swear up and down that 1,321 is the actual number of traffic deaths and not 1,837 in spite of the fact that 1,837 people really did die in traffic deaths in July.
That's how it works for employment data.
They're supposed to interview 60,000 households each month, but some Administrations have altered that figure (Clinton limited it to 40,000 households per month).
These 60,000 households are not randomly selected each month. They are pre-selected and the households agree to participate in the survey for an entire year (and yes they use cell-phones).
When they call, they will ask you if you're employed. The next battery of questions depends on your answer.
If you are employed, they'll ask how many hours a week you work. If you work less than 35 hours per week, they'll ask why you aren't:
I care for children/family/go to college/etc. That puts you here: LNU02032194
Part-time is the only job I can find. That puts you here: LNU02032196
We can't move our product or there's no business. That puts you here: LNU02032195
If your answer is "
I don't wanna" you go here: LNU02600000
If you work 35-40 hours a week, even if by your own choosing you work less than 40 hours a week, you go here: LNU02500000
People who aren't working, but want to work
and have looked for a job in the 30 days prior to the survey go here: LNU03000000
People who aren't working, but want to work,
but haven't looked for a job in the 30 days prior to the survey go here: LNU05026645
That's the Current Population Survey (CPS).
For these months:
Mar = 156,441,000
Feb = 156,167,000
----------------------
Jobs added = 274,000
Now, run 274,000 through the X-13ARIMA-SEATS software program and add in your weighting factors and voila!
You get 20,000 Seasonally Adjusted jobs added and that's what everyone seizes upon.
What about the other 254,000 workers who really do exist and are really paying taxes?
Nobody cares, because it's not "news-worthy."
A different measure that's conducted along with the CPS is the CES or Current Employment Survey.
That is a survey of private -- not government -- employers and the number of employees on the payroll at the time of the survey.
Mar = 126,994,000
Feb = 126,370,000
---------------------
Jobs added = 624,000
So, you have two different measures showing 274,000 jobs added and 624,000 jobs added but the "Seasonally Adjusted" number is statistically manipulated is "20,000."
Too bad the Department of Defense didn't use Seasonally Adjusted Casualty Figures for Vietnam, they might have gotten more support.
Instead of 274 KIA on a particular day, we could have a Seasonally Adjusted 20 KIA.
Because BLS makes adjustments to the population based on Census Bureau data each January, we need to be careful making year-round comparisons to avoid being deceitful and disingenuous.
So, August to December 2018
Dec = 156,481,000
Aug = 155,539,000
----------------------
Jobs added = 942,000
January to July 2019
Jul = 158,385,000
Jan = 154,964,000
---------------------
Jobs added 3,421,000
Total jobs added August 2018 to July 2019: 4,363,000 jobs
We can compare the CPS to the CES. Note that since the CES is based on employer payrolls and not population, there is no population adjustment in January.
Jul 2019 = 129,962,000
Aug 2018 = 127,949,000
----------------------------
Total jobs added = 2,013,000
Why the discrepancy?
It's due to the methodology, and neither is inherently flawed.
Asking employers how many people are on their payroll is going to be more accurate than asking 40,000 to 60,000 households to provide data, then comparing that data to what you believe to be is the population of the US at the time to get the data.
There are buffoons who actually believe that February's 20,000 Seasonally Adjusted jobs is the correct number and the 75,000 Seasonally Adjusted jobs in May is the correct number.
Never mind that Social Security collected these amounts in FICA payroll tax revenues:
$79,888,911,673 May
$64,407,000,000 April
-------------------
$15,481,911,673
The FICA payroll tax is 12.4% for both employer and employee.
That means $15,481,911,673 was collected from $124,854,126,395 in wages.
Right? $124,854,126,395 * 12.4% = $15,481,911,673
So, 75,000 new workers earned $1,664,721 each in the month of May.
That works out to $10,404/hour.
That's one hell of an hourly wage and those same people who are dumb enough to believe only 75,000 jobs were added in May also claim wages aren't rising.
Someone might ask,
"Why don't they just get the data from the IRS or SSA instead of wasting tax-payer money to conduct surveys?"
That's a good question and there is a very valid reason why they don't.
Federal law bars data-sharing between federal offices and agencies, and between the federal government and the States.
The IRS can get data from the Social Security Administration, they just need a search warrant based on probable cause signed by a federal judge.
The Social Security Administration can get data from the IRS, they just need a search warrant based on probable cause signed by a federal judge.
Since BLS has no law enforcement powers or administrative powers, they can never get a search warrant.
States can share data among each other and do by entering into collaborative agreements.