[bananalychee]

It's not without precedent, but the fact that initial job numbers have been consistently inflated over the last 3 years and that the magnitude of the downwards revisions is on par with 2008-2009 for two years in a row (and growing) is concerning.

[wredcoll]

How exactly were they inflated?

[ndiddy]

The monthly employment numbers the BLS publishes are basically nonsense and I'm sure they wouldn't publish them if Congress didn't force them to. At the middle of every month, they run a survey asking ~120,000 businesses how many employees they had as of the 12th of the month. It takes 8 weeks for all the responses to the survey to come in, but the initial monthly numbers are based off of the first 2 weeks of responses. This initial data is always more representative of larger companies. The BLS then generally issues two corrections to each month's data, one when all the responses to the surveys come in and the other when they get the actual unemployment numbers from quarterly unemployment insurance tax filings.

We see large corrections in employment numbers when there's rapid changes in the job market that mess with the models, or when the changes are focused towards small companies. Right-wingers have somehow decided that all of this is instead due to the BLS somehow being out to get Trump, despite there being no significant changes to how the jobs report is made since the mid-90s.

[notmyjob]

I’m not sure how the politics play out but a stat that is always off in the same direction, always over and never under, is what some statisticians call “biased.”

You can have non-biased indicators that have error with mean 0.

Maybe a better question, when judging current operations, is how precise the biased estimates are becoming overtime. Is the size of the error increasing or decreasing.

[jeffbee]

The need of downward revisions is 100% due to falling and selective response rate to the early survey.

[orwin]

If it's like in my country, it's probably because you have more and more people "self-employed", and the average "small business" went down from 3.8 employees to 2.2 over the last 6 years (made up numbers, but i've read it almost halved which caused a lot of issues).

I think we created a new status for Uber/Deliveroo and other workers to put them out of the category three years ago and it fixed a lot of our employment data issues.

[jeffbee]

You seem to be under the impression that these figures are exclusively sourced from employers, but they are not. They are sourced in part from a survey of 60,000 households every month, where each household in the the survey for several consecutive months. Here is some information about non-response rate: https://www.bls.gov/cps/methods/response_rates.htm

[logifail]

> You seem to be under the impression that these figures are exclusively sourced from employers, but they are not. They are sourced in part from a survey of 60,000 households every month, where each household in the the survey for several consecutive months

These are two separate metrics, they measure different things, and the figures often differ (unsurprisingly).

The BLS "establishment" survey (aka Current Employment Statistics, CES) surveys 120k+ businesses and government agencies, it measures jobs (not people), counting the number of payroll positions. This is "non-farm payroll employment", excluding the self-employed, farm workers, and private household workers.

The BLS "household" survey (aka Current Population Survey, CPS) surveys ~60k households, measuring individuals, whether they are employed, unemployed, or not in the labour force. These data are used to calculate the unemployment rate and labour force participation. This includes farm workers, the self-employed, and domestic workers.

[mensetmanusman]

That excuse works one time. After that, you fix incentives and hire competent data gatherers.

[logifail]

> you fix incentives and hire competent data gatherers

You assume the data gatherers were at fault... <chuckle>

These data gatherers work for their government. How do you ensure they're happy to gather and publish data which is essentially critical of that very government?

[chasd00]

I agree, if you can't publish a number that's pretty accurate then you shouldn't publish anything. That's why i wasn't broken hearted when Trump fired that BLS person. If you can't even get close then someone needs to be found that can or at least has the balls to say "idk what the number is, we're not publishing without good data".

[matthewdgreen]

Talk to Congress about this. Publishing this monthly data is a legal mandate, it's not up to the BLS head, and firing the BLS head doesn't fix it.

[mensetmanusman]

It’s almost like the leaders of those organizations were incompetent.

Such simple statistics and data gathering should be simple for a federal organization.

[spicyusername]

or its almost like gathering that data is not simple...

[logifail]

> Such simple statistics and data gathering [...]

Simple?! "Sweet summer child..."

On a more serious note, how would one ensure that a government department be sufficiently independent that it can publish data (implicitly) critical of its own political leaders without fear of retribution?

Answers on a postcard, please...

[0cf8612b2e1e]

In fact, the BLS budget has been cut (DOGEd) and they have been warning for months now that this is impacting their ability to follow up on surveys.

[logifail]

> the BLS budget has been cut (DOGEd) and they have been warning for months now that this is impacting their ability to follow up on surveys

This was broken long before DOGE was a thing:

https://seekingalpha.com/news/4142722-why-was-there-such-a-b...

"There's still ongoing chatter about the huge revision to U.S. job growth seen yesterday and what it might signify for the economy and markets. 818,000 jobs were wiped out in the 12 months through March 2024 (or 68,000 per month), resulting in the biggest downward adjustment since the global financial crisis."

[seanmcdirmid]

DOGE broke it even more. The error bars have always been non-zero. Now the error bars will be even larger than before (unless Trump just outlaws error bars).