| Submit your comments on this article |
| Home Front: Politix |
| Jobs went up only 32,000 or maybe by 629,000 depending on the survey |
| 2004-08-06 |
| I'm not sure if this is the biggest statistical discrepancy in govt. history but, if not, it is at least in the top ten. per the report from the survey of business establishments Nonfarm employment was little changed (+32,000) in July per the report from the survey of households Total employment rose by 629,000 to 139.7 million in July, and the employment population ratio--the proportion of the population age 16 and over with jobs--increased to 62.5 percent. The civilian labor force also increased over the month, rising by 577,000 to 147.9 million, and the labor force participation rate rose to 66.2 percent. It seemed to me that earlier this decade the household report was a precusor not the other way around but I haven't done a study - |
| Posted by:mhw |
| #9 mhw - how about if tax cuts generate small business jobs, and many of those are self-employed, i.e.: no W-2's, then the household survey is more accurate? |
| Posted by: Frank G 2004-08-06 7:18:56 PM |
| #8 yes Chuck but those numbers are based on the household survey, not the payroll survey |
| Posted by: mhw 2004-08-06 6:34:05 PM |
| #7 139,660,000 people were employed in July, a new record. This is an increase of 1,870,000 from January 2001 when George Bush took office and an increase of 1,094,000 in 2004 alone. This is also an increase from the bottom of the recession in January 2002 of 3,945,000. Nearly four million jobs! Data and graphs |
| Posted by: Chuck Simmins 2004-08-06 4:36:47 PM |
| #6 An individual month or few doesn't tell you anything because it could easily fall in the margin of error. damn proud, I agree with your statement above. It probably makes sense for the BLS to put a four month moving average together like what is done for the weekly new unemployment number. The only reason BLS doesn't do this is because it would be admitting exactly what your statement says. Oddly enough, this year the big gains in employment in the payroll survey have come in months with an 'r' in them. |
| Posted by: mhw 2004-08-06 2:56:36 PM |
| #5 They establishment survey is taken by I believe about 300,000 businesses in the US. It is mostly filled with the mid-size and large corps operating in the US and not so much the smaller companies. Also I'm not sure how it deals with large foreign firms that have offices in the US. The household survey is taken by calling up 60,000 households each month and asking about their employment situation. The problem with the establishment survey is that because it focuses on larger companies it doesn't reflect growth in employment when that growth is focused on small businesses or self-employed... which are both becoming more prevalent in our economy. The problem with the household survey is that since you're extrapolating up from 60,000 households to determine what's going on in 70 million households there is a larger range of error. My opinion is that the establishment survey is becoming increasingly worthless each year and the only true measurement to judge is the household survey trendline over the course of a year or so. An individual month or few doesn't tell you anything because it could easily fall in the margin of error. |
| Posted by: Damn_Proud_American 2004-08-06 2:50:30 PM |
| #4 The payroll report is the more reliable number since it surveys actual payroll headcount. The household survey numbers are softer because the survey asks how many people in your household are working. Small businesses and sole proprietorships such as consultancies are not included in the payroll survey. One weakness in the household number is that people who are severely underemployed often say that they are working. All in all, it's a mixed bag for W. Had the unemployment rate not fallen, he would be getting even more sh*t from Kerry. |
| Posted by: Tibor 2004-08-06 1:12:47 PM |
| #3 And about 30-45 days or so after the announcement they always come back and adjust the numbers. Non-Farm jobs missing from the Biz survey -- some portion are prolly self-employement or entrepreneurial jobs. |
| Posted by: .com 2004-08-06 11:18:08 AM |
| #2 AP always puts more focus on the business establishment survey, so does CNN, etc. The business establishment survey has a much bigger sample size--- so most people consider it more accurate. There were some months this year where the business survey showed more growth than the household survey. There are, IMO, two keys to understanding this: 1. there are two surveys; one is not a subset of the other; 2. our economy is very complex; people sometimes have more than one job; sometimes they even are sole propriators in more than one business; sometimes they do work with complicated contractual arrangements |
| Posted by: mhw 2004-08-06 11:12:50 AM |
| #1 I've been in this same exact argument with a co-worker all morning. Yesterday he claimed that unemployment was up yet it is actually down (MSNBC Squakbox is his favorite show, G-d knows why). Of course, when I hit him with the facts he says that new jobs had slowed, yeah that's it that's the ticket. How can they say non-farm jobs rose by only 32k when total jobs rose by a whopping 629K in July? Am I to assume that 597K FARM jobs were created? Bull. Shit. But in all seriousness, if anyone out there can make sense out of this I, for one, would appreciate it. (Good catch BTW mhw. The AP report takes this Labor information almost word-for-word, yet "forgets" all about the third paragraph. I know it's a long way down the page, so the monkeys at AP probably got tired and took a nap before reading the rest of it. They did, however, mention the unemployment figure going down. Bully for them.) |
| Posted by: Chris W. 2004-08-06 11:03:09 AM |