February 22, 2004

Department of "Huh?"

In this morning's New York Times, I read the headline "A Prettier Jobs Picture?" The headline writers are making a pun. Virginia Postrel is writing about undercounting the number of people who are at work making things (and people) prettier. The headline writers are referring to that and are also saying that this undercount means that the overall jobs picture is "prettier" than standard statistics suggest.

The headline writers are wrong. This is yet another example of something every single reporter swears: that American journalism would be much, much improved if editors would let reporters write their own headlines.

Virginia Postrel writes: "It is tempting, of course, to treat these undercounts as trivial. After all, what do 200,000 massage therapists or 300,000 manicurists matter in a country of 290 million people? But this list of occupations is hardly comprehensive. In every booming job category I looked at, official surveys were missing thousands of jobs. As the economy evolves, however, this bias against small enterprises and self-employment becomes more and more significant...." Virginia Postrel means that the detailed occupational estimates are undercounting growth of employment in aesthetic occupations. But the headline writers think she means more.

Let's go over the three different sources of data we have here:

  1. The first is the Survey of Occupations. the Survey of Occupations mails questionnaires to 200,000 businesses every six months, and uses the responses to construct estimates of employment and wages for specific occupations. This is the survey that undercounts the "200,000 massage therapists... 300,000 manicurists."

  2. The second is the Household Survey. The Household Survey knocks on the doors of 50,000 households every month, and uses the responses to construct estimates of the unemployment rate, the employment-to-population ratio, and other statistics. Household Survey interviewers ask people whether they are at work or not. The Household Survey picks up the massage therapists and manicurists (even though it does not identify them as such). In the Household Survey, moreover, there is no rise in the proportion of workers who are self-employed--there has been an uptick in the proportion of workers who say they are self-employed since its end-of-2001 low point, but the medium- and long-run trend in self-employment is down.

  3. The third is the Payroll Survey. the Payroll Survey analyzes the monthly payroll records of 400,000 businesses employing 50 million people. The Payroll Survey does not count the self-employed. The Payroll Survey does guess at the number of new businesses that have been started whose payroll records have not yet entered the system, and at least up through the middle of 2003 retrospective analyses suggest that its guesses have been rather good.

Thus there is no reason to think that the totals of nationwide employment--which are derived from these second and third of these data sources--are substantial undercounts because of any significant "bias against small enterprises and self-employment."

Essay: A Prettier Jobs Picture?: By VIRGINIA POSTREL: ...The official job counters at the Bureau of Labor Statistics don't do much to overcome our blind spots. The bureau is good at counting people who work for large organizations in well-defined, long-established occupations. It is much less adept at counting employees in small businesses, simply because there are too many small enterprises to representatively sample them. The bureau's occupational survey, which might suggest which jobs are growing, doesn't count self-employed people or partners in unincorporated businesses at all. And many of today's growing industries, the ones adding jobs even amid the recession, are comprised largely of small companies and self-employed individuals. That is particularly true for aesthetic crafts, from graphic designers and cosmetic dentists to gardeners. These specialists' skills are in ever greater demand, yet they tend to work for themselves or in partnerships.

Consider the ubiquitous granite counter top. The slabs are imported, but the counter tops are made in the United States, and the shops that do the work are proliferating rapidly. ''It's an explosive trend,'' says Michael Reis, editor of the industry magazine Stone World. In 2002 alone, the magazine added 2,000 fabricators to its 20,000 subscribers. Reis estimates that there are 8,000 to 10,000 fabrication shops in the country. Equipping a fabricating business can cost less than $30,000, thanks to relatively inexpensive machinery developed over the past decade.... The fabricators are small local shops, usually with fewer than 10 employees, which makes stone fabrication the sort of industry job prognostications tend to overlook. The Bureau of Labor Statistics survey doesn't have a separate category for this burgeoning construction craft. It is lumped under ''cutting and slicing machine setters, operators, and tenders,'' a production category dominated by people slicing paper in mills and printing plants. So losses in those fields mask the growth among stone crafters.

Or take Denise Revely.... She used to work in an electronics factory. Now she gives facials. She used to draw a paycheck from a day spa. Now she works for herself. Her business has been in the black ever since she opened last July. But as far as official statistics are concerned, she doesn't exist....

It is tempting, of course, to treat these undercounts as trivial. After all, what do 200,000 massage therapists or 300,000 manicurists matter in a country of 290 million people? But this list of occupations is hardly comprehensive. In every booming job category I looked at, official surveys were missing thousands of jobs. As the economy evolves, however, this bias against small enterprises and self-employment becomes more and more significant....

Posted by DeLong at February 22, 2004 09:46 AM | TrackBack | | Other weblogs commenting on this post

How accurate is the household survey? If they knock on doors, suppose there is no answer. How do they follow that up? Or suppose the resident refuses to divulge any information? It would be nice to know the rate of non-response and how they do imputation.

Posted by: A. Zarkov on February 22, 2004 09:59 AM


This won't work. The GOP is trying to browbeat people into believing that the economy is better than what their own experience tells them. This will not have the desired effect of convincing the legions of voters that the economy is fine. What it can do is mislead the political elite into believing that their policies are more effect or better than reality suggests.

IL is looking to be a blowout for the Dems. They will probably pick up a senate seat there. IN will keep their senate seat and OH is set to abandon Bush. 3 times an hour the Chicago radio ads drone on and on about the number of layoffs, plant closings and mfg job losses. These are resonating with the voters and accurately express the job market they experience. Taking a job for $10 per h and no benefits in a massage parlor or beauty shop is not appealing to workers who cannot replace their $20/h plus benefits mfg jobs that are lost. The job loss is being addressed by pouring money into new mfg technology and nanotech mfg, but that will not blossom until long after the 04 election. Mr Bush has been MIA for the last 3 years in supporting this transition or trying to speed its implementation. It is not even on his radar yet. His one jobs program, oil drilling jobs in Alaska are not what Midwestern voters have in mind.

Posted by: bakho on February 22, 2004 10:02 AM


Jeez, are we to be happy about all those nail
shops in the malls where new immigrants work
for the minimum wage with not benefits??

Posted by: Bartolo on February 22, 2004 10:08 AM


Virginia is making a good point. There are all kinds of independents in construction and repair occupations. Last fall a heavy rainfall resulted in water pouring into my house through a leaky skylight. After putting a tarp over it I then went through the frustration of trying to find someone to fix it.

I'm about a 40 minute drive outside Seattle (where the unemployment rate is supposedly over 7%), in a semi-rural area. Just as I was about to give in to paying $3,000--the minimum economical amount for any roofing company--to a big roofing company, I got a call returning a message I'd left on an answering machine. It was a one man band working out of his home and pick-up. I had a new skylight the next day for $500.

His business isn't being counted by anyone (he's in a semi-rural area too). My business isn't being counted, nor is the landscaper who rents home and office space from me, nor any of the myriad dump truck owners, backhoe operators, car and truck repairmen, hair stylists, welders, tree surgeons, carpenters, septic tank servicers, parking lot stripers, and on and on, who advertise themselves with handmade signs posted on trees and telephone poles.

Nor are the pianist, guitarist, and sax player I'm listening to practice as I type. It's not an insignificant number of jobs. From what I observe of parking lots in malls and waiting lines to get into restaurants around here, I'd say Ray Fair is correct. Bush will have a relatively easy win come November. And it won't be because any GOP consultant brainwashed everyone.

Posted by: Patrick R. Sullivan on February 22, 2004 10:41 AM


Brad, It seems to me the headline writers understand V. Postel's meaning better than you. My conclusions are based only on the quoted portion of the article.
She writes: "It is tempting, of course, to treat these undercounts as trivial." Implication: The undercounts are NOT trivial. They matter to the employment picture.

Further: "In every booming job category I looked at, official surveys were missing thousands of jobs." I read that to say the official surveys (ALL of them) are substantially undercounting jobs.

The statement "This [undercounting] is particularly true for aesthetic crafts..." Hardly seems the main point of the article.


Posted by: Jeff on February 22, 2004 10:46 AM


Yeah, this is gonna work. Don't like the numbers? No problem, we'll get you some different numbers. Fuckwits!

Posted by: SW on February 22, 2004 10:51 AM


> His business isn't being counted by anyone

Probably including the IRS.

Shorter Patrick Sullivan: I support cowboy roofers.

Posted by: blah on February 22, 2004 11:15 AM


I would presume, even if Postrel were correct, that secondary effects of all those uncounted jobs would show up at McDonalds and Ford in the usual counted ways.

Posted by: bob mcmanus on February 22, 2004 11:18 AM


The question is, should burger-flippers be counted as "aesthetic crafts" jobs? Or maybe just Subway "sandwich artists."

Posted by: Grumpy on February 22, 2004 11:22 AM


As long as Ginny Postrel is working well and happily, there will be no jobs problem in America. Decent wages? Health care benefits? Who need 'em?

Posted by: lise on February 22, 2004 11:33 AM


Does dipping the fries count as IT employment?

Posted by: lise on February 22, 2004 11:35 AM


The reason this article is significant is that the payroll survey is the number being used consistently by those complaining that we aren't getting enough new jobs. I would bet that most of the public is not even aware that there are two numbers.

If you look at the actual BLS report (January is here:)


non-farm employment (payroll) went up 112,000. This "paltry" number has been in every headline in the country. Household employment went up by almost half a million. How many headlines did that show up in?

It makes a big difference what numbers you look at, and this needs to be mentioned when talking about these things.

This raises the mystery of why there is such a discrepancy between household and payroll employment. A pie chart I saw somewhere (can't find the link, dammit) showed that self-employment counted for only about a quarter of these jobs, and the article the graph was in had no real idea what the rest were.

Posted by: tbrosz on February 22, 2004 11:42 AM


Shorter Patrick Sullivan:

"How can a survey be right if it didn't count me!"

By the way, did people only advertise for odd-jobs starting in 2001? If not, how do odd-jobs explain the recent downturn/discrepancy? Isn't the population overestimation the Occam's Razor explanation? Or do you know more than Alan Greenspan does?

Posted by: strawman on February 22, 2004 11:45 AM


Important caveat: Note that the household employment was subject to a statistical hiccup based on an adjustment made to the way they do the numbers (I've read the explanatory article through twice and still don't get it-


-The "befores" and "afters" don't seem to match up with the published numbers.) So the January household numbers may be misleading. Once this has worked through the system, we'll get better numbers.

Posted by: tbrosz on February 22, 2004 11:50 AM


Postrel's arguments that the BLS undercounts jobs in professions such as stone cutting (granite kitchen counter tops) and manicurists and massage therapists and gardeners and graphic designers, as well as many other categories, may be right.

The problem is that the BLS has always under counted these jobs. Always.

What is missing from the advocates of the Household survey is a persuasive explanation as to why this is a new phenomena. Show me that these are brand new jobs, created during the past 3 years.

Consider the tortured path we must take to rely upon that conclusion: anecdotal evidence of historically under counted jobs, which we then must believe on faith are being created anew. This, despite the Fed chair saying this data source (household survey) is not to be strongly relied upon.

Not only must we use the "data must be bad" cliche to prove the point -- but in this case, we must also ignore the data which the Fed notes as "more accurate"

OK, rely on the bad stuff, discard the good stuff, overly rely upon anecdotal evidence, and make several unsupported leaps of faith. Forgive my skepticism, but I tend to get suspicious when a flagellating argument runs counter to everyday ordinary experiences.

Pretty Job Picture? Household survey advocates need to take off their beer goggles . . .

Posted by: Barry Ritholtz on February 22, 2004 11:55 AM



at dinner the other night, a woman at a table next to us was loudly discussing her husband's plight. Apparently, despite being an independent contractor and having been with the company for 25 years, he needed to get out from under a lousy boss, who had only been there 4.

Now, my understanding is that, like medieval judges distinguishing between serfs and free peasants, the IRS takes a dim view of "independent" contractors who are not only told what to do, but how to do it, so I'm not sure how bad a boss could really be. In any case, it seems really unclear on the concept to say a (nominally independent) contractor is "with the company".

Could the difference between household and payroll surveys lie in the payroll people generally being more accurate about affiliation?

Posted by: Dave on February 22, 2004 12:00 PM



The Bureau of Labor Statistics has emphasized that the payroll survey is far more accurate than the household survey. Either way there is a job creation and wage and benefits problem of significant proportions.

"Nevertheless, regardless of which survey is used, the labor market is still characterized by a significant jobs deficit, and this remains the weakest jobs recovery on record. Outside of retail and accounting for returning strikers, net hiring amounted to 27,000, far below the 150,000 benchmark of that is needed to keep up with the growth of the working-age population. The lack of sustained job creation at a level commensurate with this stage of the economic expansion is generating long-term joblessness and stagnant wage growth."

Posted by: anne on February 22, 2004 12:18 PM


If they knock on doors, suppose there is no answer. How do they follow that up? Or suppose the resident refuses to divulge any information?

My SO worked for the Household Survey in the early 80's--a tough time here in the midwest. There was a great deal of pressure (loss of job) on the interviewers to complete all interviews. She would track people down, even to the extent of interviewing them in bars. In one instance, she tracked down an interviewee who was on summer National Guard duty in the Texas desert and interviewed him by phone.

She couldn't give up after one failed attempt, but had to try everything possible to get the interview. If person was uncooperative, he would be contacted by a supervisor. The Bureau tried to hire only women as interviewers, as it was thought people would be more likely to cooperate with a woman.

She would often put in 50-60 hours during interview week. No overtime and $5.00 per hour.

Whether these conditions persist today, I can't say.

Posted by: Tom on February 22, 2004 12:41 PM


Even if it's true that jobs that would be hard to count with the payroll survey are booming (and I won't take Postrel's word for it), I assume manicurists and cosmetic dentists have doors which can be knocked on by household employment surveyors. And the statistics show that Postrel is making a mountain out of a molehill, at best, and noxiously riding the horse of dissimulation at worst.

Posted by: rps on February 22, 2004 01:31 PM


" did people only advertise for odd-jobs starting in 2001?"

Something like that, yeah. When you lose a job with a corporation one quick way to earn money is to start your own business. And when those corporations are slow to rehire more people stay self-employed.

In the specific case of my roofer, it had been only a part-time gig until he was laid off from a trucking firm he was driving for. So he made the roofing a full-time gig.

Posted by: Patrick R. Sullivan on February 22, 2004 01:36 PM


I don't know how Ray Fair predicts the states, but applying national numbers can be misleading. Job loss in the Midwest started first and continues even now. If you look at the swing states that Brad posted above, note how many Great Lakes mfg states are relatively closely contested. MI, MN PA WI and OH are all in this group and are a lot of electoral votes. Of the Great Lakes states, only IL and NY are comfortably in the DEM column and IN in the GOP column.

Bush will have a tough sell in any of these states. The job loss here is very high, moreso than on the coasts.

Posted by: bakho on February 22, 2004 01:37 PM


Fuck em all! We are about to enter a period that not one economist has ever predicted. Between free trade, higher productivity, (Think computer age) and outsourcing we are on a path that not one person has had the foresite to see. We are entering an age in America where there will be no way by any means to employ more than 75% of the people who want to work. If a country does not produce any of it's own goods(foodstuff's not included)there is no way possible to sustain the way of life that we know. There is no way humanly possible that a completely service based economy can sustain the American way of life. Now I am not an economist, but I assure you that if these economists begin to do the math, they will see that when you go into any store, and every product is not made in America, that what I say is the truth. Be very afraid, because if things continue this way,we are about to head on a path towards anarchy.

Posted by: John Jason on February 22, 2004 02:14 PM


She is simply playing games with numbers.

One-- employment by ocupation comes from the
payroll data not the household survey.

two -- the household survey always shows more
employment than the payroll ddat.

Household employment as a % of payroll
employment was about 135% in the 1970s
began a gradual long term trend about
1980 that bottomed at 122% in 2000
since rebounded to 127.5 % this january.

Unless she is trying to make the case that
payroll employment is undercounting
actual employment by more that 25% to 33% her
mumbers do not mean anything.

In the household survey they do not ask
what is your job. On this basis the number
of extra jobs she implies that the government is not counting are in the household survey.

Her article is just a bunch of numbers strung together that do not mean anything.

Posted by: spencer on February 22, 2004 02:29 PM


But surely people have lost jobs in recessions before, right? So what's different about this one that would cause them to be undercounted? Seems more like Occam's dull edged butter knife to me

Posted by: strawman on February 22, 2004 02:37 PM


i get tired of the comments that they did not ask me so the household survey is useless type of comment.

It is as good a statistical survey as any that exist.

Almost all large business depend heavily on statistical sample type survies that are no better
than this one. It has a over a 50 year
track record and no serious study has ever found a serious flaw in it.

Posted by: spencer on February 22, 2004 02:38 PM


Shorter John Jason

"I am no economist, just Nostradamus"

Posted by: strawman on February 22, 2004 02:39 PM


Did most of you even read Brad's post.


Posted by: Atrios on February 22, 2004 03:21 PM


Patrick R. Sullivan: survey coverage

What makes you think that your roof guy is not covered? If he has a business license and writes you a proposal/estimate with a description of the work and a bill, on which you can get him on the hook in case of a defective job, then this is legal business, and should be probabilistically covered by the survey. (I would presume the surveyors select establishments from the county/state business registries.)

If the guy does the work under the table (business license or not), essentially like a craft-skilled friend driving up to your house, only that you have to pay for the labor, then sorry, you can't ask the surveyors to count people performing unreported jobs in the shadow economy. You can't have it both ways -- people not reporting their activities to the DOL/IRS/FICA, but reporting to the employment surveys.

(And hopefully, if it turns out the job was not proper, the guy will agree to a warranty repair in either case.)

Posted by: cm on February 22, 2004 04:08 PM


Dr. deLong:

Thanks for addressing this issue. You said what I though you'd say, but it's nice to have it in writing. Being a non-economist, I worry about messing up inferences in the area.

Posted by: Jonathan Goldberg on February 22, 2004 04:26 PM


You nailed it once again Atrios! Read the "blog".
Postrel is a GOP anecdotal groupie.

The real questions to ask us newly self employed "captains of industry" are:
1.Did you become self employed as a result of failure tofind a job.
2. Are you making as good a living as when you were working at your real (career) job?

Posted by: dlbeal on February 22, 2004 04:55 PM



Ray Fair doesn't predict electoral votes. He predicts the share of the two party vote. In 2000 he predicted Gore to get 50.8% of that, and Gore got 50.3%. That was better than any of the polls, conducted in the last week before the election.

His formula is eminently sensible. Going back to 1916, the only election it failed was 1992. When there were a couple of oddities; the primary challenge to Bush by Buchannan, Perot's fabulously self-funded third party challenge, and an exceptionally poor performance by Bush in the debates.


The roofer won't get surveyed because he's located in a semi-rural area. I'm on 26 acres, most people are on at least 5 acres. It would be impossible to "knock on doors" here.

Driving one mile in either direction on the highway from where I'm sitting, I can think of four construction businesses, a shop that rebuilds 16 wheelers that have been totaled in accidents, a Christmas tree farm, a hair salon, a bookkeeping service and a welder.

Posted by: Patrick R. Sullivan on February 22, 2004 05:34 PM


more on Virgina Postrel:

I am a memeber of 3 professional organizations.
the american association of economists(AEA)
the national association of bisiness economists(NABE)
and the american association for investment research and management(AIMR).

Moreover, my wife is a memeber of the Va. Nurses
Association even though she has not lived in Va for 30 years or worked as a nurse for 35 years.
but she has to keep her membership active to continue being a Registered Nurse so if she ever wants to go back to being a RN she does not have to pass the exam again.

In the AIMR latest directory 1.1% of the members list their job title as economists. but I know many portfolio managers,directors of research
and investment strategists that also belong to NABE. I am willing to bet that almost every AIMR member that calls themselve an economist belongs to at least NABE and probably AEA.
Moreover at meetings of the Boston Association of Business Economists(BABE) -- you have to be a memeber of NABE to belong to BABE -- about half of those attending are not primarily economists-- they are business planers, professional investors, corp. treasures, etc.

these are all just personal exambles that memeber-ship in an association has little or no bearing on whether or not you work in that profession.
Moreover, when you start looking at memebership in professional organizations the first thing you see is that there is a problem of double membership. Memership in professional organizations is a circle of overlapping circles.

Unless Postrel is willing to make some type of adjustment for this overlapping memership & inactive membership _-- as the example of my wife in the Va. Nurses Association -- I will repeat my earlier post that her article is just a listing of meaningless numbers that has nothing to do with the validity of the national employment data.

Interestingly, I just did a GOGLE of American ASSociation of * , and national Association of *.
I got 2,180,000 hits for American Association of
* and 2,960,000 hits for national Association of
*. In each one I went to page 100 of the listing without seeing any duplication. On page 100 of the American Associaton of * the American Economic Association was listed. on page 100 of
national Association of * the National Association of EState Agents was listed.

If the hits of about 6 million is accurate, that means there is an american Association of * or
national ASsociation of * for every 5 americans.
Wurely there has to be a major problem of double counting.

Posted by: spencer on February 22, 2004 05:36 PM


" Household employment as a % of payroll
employment was about 135% in the 1970s
began a gradual long term trend about
1980 that bottomed at 122% in 2000
since rebounded to 127.5 % this january."

Which would tend to support Virginia's contention.

Posted by: Patrick R. Sullivan on February 22, 2004 05:36 PM


Patrick R. Sullivan: "The roofer won't get surveyed because he's located in a semi-rural area. I'm on 26 acres, most people are on at least 5 acres. It would be impossible to "knock on doors" here."

How does the semi-rural thing relate to business surveying? Are you suggesting that the survey covers only metro areas? And of course the "knocking on doors" is metaphorical. Come on. When you register a business or participate in the census, you have to give a contact address and phone number, at least in our county.

How do you know the methodology of the establishment survey? Can you provide a link?

Posted by: cm on February 22, 2004 06:32 PM


dlbeal: "1.Did you become self employed as a result of failure tofind a job. 2. Are you making as good a living as when you were working at your real (career) job?"

I know one case in some detail, and that answer would be (1) yes, (2) no (after considering missing health benefits and stock options).

I suspect something similar would apply for quite a large proportion of the newly self-"employed".

Posted by: cm on February 22, 2004 06:42 PM


You can tell when a group has left minority status and joined the mainstream: status is not just for the competent. Virginia Postrel, Elisabeth Bumiller, Ceci Connally, and Condoleeza Rice are just a few examples of the success of Women in America.

Posted by: masaccio on February 22, 2004 07:34 PM


"We have a lot of people who will do what they have to in order to make a buck during a downturn...this could be thought of as a boom in entrpreneurial activity or a boom in desperation during a jobless recovery."

Source: Economists for Dean

Posted by: bhaim on February 22, 2004 08:00 PM


Well having worked for 20 years, 14 with the DoD I have never been contacted by the household survey. So assuming that there has always been 285million in the US I would have had a 1 in 24 chance of being polled, never happened. So anyone who has 20 years of work should ask 24 people if they have been polled. See what you find. I have no idea what the polling standard is but this would help you understand where these numbers come from.

As for P Sullivan I can only assume he is naturally opposed to opinions different from his, everyone else must be wrong.

Living in Silicon Valley and having 20 years as an engineer I can say that the household survey is wrong. This is from not only personal experience but anecdotal accounts from former unemployed coworkers.

Doing construction work for friends does not make me employed, even when paid. With no jobs available I've been living off savings.

So to those like greenspan saying you have to be retrained or others saying you aren't looking you are full of crap. The job picture looks worse than 1983 when I graduated and spent 14 months bartending waiting for my first engineering job.

Posted by: RC on February 22, 2004 08:10 PM


RC: Your nickname suggests you may be an electrical engineer, but then perhaps not.

Don't forget that the household survey is nationwide, and the Bay Area is affected differently (and disproportionately as engineers are concerned). That it does not agree with what you see in your (our) region does not make it wrong.

When I look at all those Available signs and deserted parking lots in the Bay Area's industrial park ghost towns, it's scary.

Posted by: cm on February 22, 2004 08:28 PM



The first 11 years were in boston during the reagan recession. My point is that those paying the salaries along with the significance 400,000 businesses vs. 50,000 households lead me to believe the payroll survey. My anecdotal evidence supports the payroll survey.

As a trained aerospace engineer working first in Air Force missiles, Air Force radars, network test equipment(satartup), then telecom equipment(startup) I have to say this time is different. Of course this is only personal experience but quite a bit of it.

Posted by: rc on February 22, 2004 08:50 PM


Click my url if you want to see my opinion of Virginia's column.

One point I think needs to be emphasized (which I didn't address in my blog): Health insurance.

I know this is OT, but I think it is wrong that health insurance is purchased through your employer. I don't buy my car insurance or house insurance through my employer. By having to purchase my health insurance through my employer, I am de facto giving them an edge at the negotiating table (i.e., I know people who stay with an employer because their spouse has pre-existing conditions, and wouldn't be able to obtain affordable health insurance. Thus, the employee is relegated to accepting lower wages/benefits.)

Posted by: cj on February 23, 2004 01:03 AM


The Bureau of Labor Statistics has emphasized that the payroll survey is far more accurate than the household survey.

And pardon me for asking, but how the hell could they possibly know that? Did they count all the employed people first, and then compare the results? The payroll survey, as designed, can't count newly created businesses or the self-employed. So the BLS adjusts for that. Are those adjustments accurate? Again, how the hell can they possibly know that? Even if they somehow know that the adjustments were valid in the past, how can they know if the adjustments are still accurate?

Posted by: David Nieporent on February 23, 2004 02:23 AM


Pat Sullivan-- would a decline in the spread between the household survey and payroll survey
suggest that fewer people were in small
business or self employment & that more were working in larger organizations. This is just the opposite point VP is making.

In evaluating the BLS data the point to remember is that it is rebased to the data collected in the census every 10 years. Both sets of BLS data are really based on the census data that is about as close to a 100% survey as we can get.

Posted by: spencer on February 23, 2004 06:31 AM


VP complains about an alleged bias in the reporting of employment. So what does she do? She highlights the possible places where job gains have not been counted fully but does not discuss the possible places where job losses have not been counted fullly. Which is truly a biased approach.

Posted by: Harold McClure on February 23, 2004 06:33 AM


Shorter Postrel:

"The servant class is booming. That means the economy must be getting better."

Posted by: Stirling Newberry on February 23, 2004 07:10 AM


The servant class is booming?

I thought her point was the ARTIST class is booming.

Even stipulating the "service" point, I see a higher quality of "servant" evolving. For instance: another industry she does NOT mention is the elective medical procedure. Liposuction. Lasik vision correction. Tooth whitening. Non-invasive endoscopy. General MRI. "Sport medicine" practices. "Sex-for-Life" clinics. These are doctors and technicians offering services not "necessary" to health and not paid for by insurance or medicare -- but fill (create) a demand. (And I leave as a separate category the vast numbers of doctor-lawyer combination practices that ONLY want to "help" victims of auto and on-the-job accidents.) We're talking "servants" with advanced creditials as doctors, nurses, radiologists, physical therapists, -- or at least chiropractors and "aroma chemists".

These are not shoeshine boys with wooden boxes hoping for nickles from the guys waiting for trains. (Not that there's anything wrong with that job, either...)

Posted by: Pouncer on February 23, 2004 09:20 AM


"As the economy evolves, however, this bias against small enterprises and self-employment becomes more and more significant."

For those that keep saying "but nothing's changed". Virginia is saying it has. Buy it or not, but don't keep on repeating that undercounts in the past = undercounts now unless you address her reasoning.

Posted by: Matthew Ryan on February 23, 2004 01:27 PM


Regarding the question on 'freedom' fries and the posts on Subway, McDonald's etc. Everyone knows those are a combination of culinary arts, chemical engineering and nutrition coach. C'mon people get your industrial organization categories right; doesn't anyone read Jamie Galbraith anymore?!

Posted by: Diodorus on February 23, 2004 05:52 PM


"Nail artist" is to "artist" as "house painter" is to "artist". It's a service, and a precarious existence. It may be better than nothing but it's nothing to write home about either.

Posted by: Steve Cohen on February 23, 2004 06:06 PM


And as far as gov revenue is concerned, these new "artists" are not contributing as much as they did when they were working as payrolled artisans, no? So those who want to use the household employment numbers instead of the payroll numbers are skating on thin ice. The shift to non-payroll income means less revenue for the gov coffers--not exactly the tax cut BushCo had in mind.
I have to 'fess up that the plastic surgeons did not come to mind with the expression "self-employed". Seems to me that this bunch is not only pay-rolled but pay-rollers for secretaries and receptionists and...Or are you just pulling my leg? (as a part-time, self-employed
therapist?) Is there such a thing as a self-employed dentist? yes, and stretched ones at that--doing morning newspaper deliveries. Sure.

Posted by: calmo on February 24, 2004 12:07 AM


Reminds me of this Herbert Hoover quote from his memoirs:

"Many persons left their jobs for the more profitable one of selling apples."

Posted by: praktike on February 24, 2004 12:25 PM


Tom "National Guard duty in the Texas desert"
Hey this thread is supposed to be about employment statistics not the national Guard in Texas (plenty on that in other threads).

I notice very little discussion of fact that if someone has two jobs he is counted twice by the employer survey. I would say that the gap is getting back to normal (roughly equal to number of self employed). The anomaly is that the series were so close in 2000. I (http://rjwaldmann.blogspot.com) would guess that a lot of the jobs that have been lost were second jobs (seems to fit very hot economy cooling off and staying cool no ?).

Posted by: Robert Waldmann on February 25, 2004 06:23 AM


It's not all lies - not all of it. That's the age-old dilemma.

Posted by: Bromberg Brian on March 17, 2004 11:30 PM


Reality is not affected by our apprehension of it.

Posted by: Novins Lisa on May 2, 2004 03:11 PM


Of all the words of mice and men, the saddest are `It might have been.

Posted by: Landolf Diane Wright on May 3, 2004 02:21 AM


A person never tells you anything until contradicted.

Posted by: Bailey Jake on May 20, 2004 10:03 AM


Nice site. Keep up the good work.

Posted by: discount phentermine on June 6, 2004 08:35 PM


Gratitude is merely the secret hope of further favors.

Posted by: Perlov George on June 30, 2004 07:01 AM


Post a comment