To me “conducts the survey” means collects and processes the data.
For the CPS, the raw data goes straight to BLS from the interviewer.
BLS does not take the results of the survey from Census, they take the raw
data straight from the interviewer. Census doesn’t do any of the data processing for the BLS parts.
So eho the Census director reports to is irrelevant.
I didn’t say the Census Bureau ‘processes’ the data. By ‘conducts the survey’, I mean exactly that - Census workers actually collecting the data by carrying out interviews.
I’m curious what you mean when you say the BLS gets the results directly from the interviewer. Are you saying the Census workers upload their individual data directly to the BLS? I would assume that they would collect their data and the Census Bureau would maintain the database and then transmit the raw data to BLS once it’s been collected. No, they don’t do any processing on the numbers - they give the raw data to BLS. But if they collect the data and package it for BLS, there is still a need for oversight to ensure data integrity. The obvious way that the data would be ‘cooked’ would be to subtly change the raw data, or for politically-motivated census takers to change the answers they get from an interviewee. There are ways to prevent that kind of fraud and abuse, but those techniques would have to be used by whoever receives the raw data and puts it into a database, and by the organization that actually hires the people who poll the public.
Of course, this is all theoretical since you don’t actually think the numbers were manipulated, right?
I think you are on very solid ground here, Sam, indeed, that is the obvious way. And, just as you say, there are ways to prevent this sort of thing from happening. However, those techniques would have to be implemented for the desired effect to be realized.
I see no point that I can argue with, I fold.
Gallup’s unemployment numbers for the last 30 days is 7.5%. I’m guessing they get them by calling people and asking them if they’re working (they also have numbers for part-time employment, etc). Seems to be within the same margin as the official numbers.
No, I don’t. I’m just defending my post from people trying to nit-pick it apart.
The number that is more interesting is ‘U6’, which counts people who have stopped looking for work, or who are working part-time when they would rather be full-time. It provides a better snapshot of what’s actually going on in the jobs market.
One of the anomalies of this report is that U6 remained unchanged at 14.7%, while U3 (the 7.8 number) declined substantially.
Here’s one way this could happen: Some large retail chains have said that because of Obamacare they are going to have to cut back on full-time employees and hire more part-time employees. So if KFC fires 10 full-time employees and replaces them with 20 part-time employees, U6 would remain unchanged and U3 would decline.
That may in fact be part of what’s happening, since the big gains in the household survey were part-time jobs between people 18-24 years of age. They would be exactly the kind of people who might have their full-time retail or service job turned into a part-time job.
The higher percentage of part-time workers theory is perfectly valid with respect to the overall economy of the last three years, though the better evidence is not Obamacare as the cause but rather sticky wages.
This particular gainwas mostly not part-time jobs, however.
I have no qualms with folks trying to get underneath the numbers and trying to decipher them. IN fact there is evidence that the unemployment rate was overestimated in previous months, but nothing systemic. The real issue in term of the political implications is that Romney had been touting “over 8%” for months now, so if the BLS numbers are good enough then, they should be good enough now.
What’s interesting to me is that two separate entities have polled unemployment and found it to be below 8%
I get that people want to say that it’s not important for this or that reason, that wasn’t my point. My point was to show that if the numbers have been cooked, they’ve been cooked to line up with Gallup’s numbers. Or both the BLS and Gallup are in cahoots. Neither seems likely to me.
[/rant]
I find it difficult to believe that people are surprised by the numbers.
SCHOOL STARTED. All of the major gains are related to this and the support [many now privatized] organizations related to education.
Gee, let’s see…bus drivers are up, food service is up, ambulatory health care and social assistance is up, state and local education is up. Hmmm, I wonder why that occurred? Could there possibly be something that happens in September to affect this? Autumnal equinox? Dancing with the Stars premiere? NFL Season starts? Oh, yeah–that school thing.
Look at Table B-1, http://www.bls.gov/news.release/empsit.t17.htm
Get underneath the numbers, because the punditry at neither FOX nor MSNBC will do it for you.
If you are so foolish as to think that “Obama added X00,00 jobs” or alternatively “the BLS cooked the books” then you deserve to no respect by anyone for parroting either of these positions.
[/end rant]
So, there’s a measurable drop in the unemployment rate every September?
The top-level data in the BLS report is seasonally-adjusted, so it should account for that.
The only way the start of the school year should have an effect is if students this year are behaving differently than they have in the past. For example, I seem to recall that the year after the crash there was an anomaly in the data because more people than usual went back to school because of the job situation.
Another thing that could do it would be an uptick in students taking part-time jobs due to the increasing cost of education. In many areas, there are part-time jobs that go unfilled even when the unemployment rate is fairly high - for example, evening clerks at a convenience store. If more students than usual were taking such jobs, it would show as an increase in the job stats.
I’m not saying that’s the case - just an example of how seasonal adjustment may not reflect the situation in this particular season. I have no reason to believe that this specific example is true.
Surely it’s because business are anticipating a Romney presidency right? Notice how the rate drops as Romney’s poll numbers improve?
And yet today’s numbers on initial unemployment claims show a sharp increase from last week, and significantly higher than expected.
Just call Jack Welch. He can explain it!
Your same cite clearly shows the number of insured claims down significantly, which taken all data in total gives the net decrease in the unemployment rate.