Email a link to this entry to:


Your email address:


Message (optional):

Over the last decade the labour force size has been increasing faster than the rate of population increase of the over 16 segment. Since 2000 the size of the seasonally adjusted labour force has held basically steady while the unemployment rate has risen and then steadied over the last few months. In the 90's the reason 1.5% increases basically held unemployment rates steady seems to be because of the steady increase in the labour force. Since the labor force size is no longer increasing one wouldn't, I think, expect the relationship to hold. However that begs the question of why didn't the discrepancy show up in 2000 when the labor force stopped growing?

Perhaps it has something to do with seasonal adjustment - I look at the graph from 92 of the labor force size seasonally adjusted and non-seasonally adjusted and I see the first graph stalling out in 2000. I see the second graph basically continuing trend - there is a slight decrease in the participation rate, but the gross numbers basically continue to rise.

What exactly this means I leave to the experts to decide - but at a glance it looks like something is wrong with the seasonal adjustment. I'd expect numbers between seasonally adjusted and non seasonally adjusted labour force numbers to vary - but I would expect both charts to trend together.

Probably I'm missing something obvious, but this has me scratching my head.

Posted by: Ian Welsh on November 1, 2002 09:41 PM

Please ignore the previous. It is incorrect. My apologies for the error. The two graphs do indeed move together and I'm not sure what graphs I was comparing.

Posted by: Ian Welsh on November 1, 2002 10:31 PM

>>The source of this mystery does not lie in any statistical discrepancy between the employer survey from which we calculate payroll unemployment and the household survey from which we calculate the unemployment rate.<<

• Well if the current household survey numbers are adjusted the same way as the previous ones then I'm still not fully convinced.

• I have already mentionned that people seem to have bravely taken on part-time jobs. One of the things that has striken me in California is the number of people who (used to) work several jobs. These days, some have retrenched to a single part-time job. Yet, this does not affect the estimated unemployment rate in any way. It is nonetheless a great feature of the aclaimed flexibility of the American labor market.

• As you wrote, people do fall off the labor force statistics if they declare they haven't looked for a job for the last week, if I'm correct. That's pretty drastic in a morbib labor market.

• One last thing: are there some demographic changes at work here? People born in 1945 are turning 58 years old. Isn't that the earliest retirement age for entitlement to Social Security in the US? When did the baby boom really start?
  

Posted by: Jean-Philippe Stijns on November 2, 2002 12:24 AM

Since the unemployment rate has not been climbing in recent months, could it be that the lack of job creation is not as serious a problem as it appears?

Still, the lack of job creation has to be putting pressure on many families even if it just means incomes have been rising more slowly.

Also, note that African-American unemployment is approaching 10% after years decisively below that level.

How serious then is the lack of creation?

Posted by: on November 2, 2002 01:43 PM

lack of unemployment jump is simple to explain.

this has been a long recession... more long term unemployment after 6 months of unemployment insurance these unemployed aren't counted. In other words there hasn't been any black monday shake up in economy or some sort of 'crash" rather a slow steady drop in jobs .. layoffs have been consistent the last two years. so literally as one group of people fall off the labor statistics then another group replace them.

in the gutter is about 3% more unemployed who are burning thru savings, going to limit on credit cards and praying for an extension in unemployment when democrats take over congress in november

Posted by: smartone on November 2, 2002 01:52 PM

Recent economic data seem to add strongly to the deflation threat that DeLong, Krugman, and Roach have been worried about.

If the Fed is going to begin lowering interest rates again, I imagine the governors are coming to the deflation worry as well. Japan is in deflation and slowing in growth again. China is in deflation, but growing well. Deflation has begaun in several other Asian economies. Germany is slowing again, and there is little price pressure in the economy though wages are strong.

Stephen Roach may have been especially prescient about American post-bubble problems rippling about the world.

Posted by: on November 2, 2002 02:32 PM

Not since 1980-1981 have so many jobs been lost as in the current 2 year period. Still, there seems less worry about the losses. Curious.

The continued cuts in manufacturing jobs takes us back to 1961 in manufacturing employment levels. Service and construction jobs are what we must now depend on.

Posted by: on November 2, 2002 02:41 PM

'Which is a better description of the labor market in 2002? That the unemployment rate has been flat? Or that payroll growth has been so slow that one would have expected the unemployment rate to rise by more than three-quarters of a percentage point?'

That all the action all those people with tenous connections to the labor market, who finally joined it in the 1990s, moving back off the statistics? I have no idea how to verify this theory, unfortunately. Local anecdotal evidence supports it, though.

Posted by: Jason McCullough on November 2, 2002 03:28 PM

I think Ian Welsh may have had the right idea in his opening lines as to the the solution to this.

It seems the lower-than-would-be-expected unemployment rate may be explained by a clear break in the BLS data in 2002 -- following another break that went in the other direction around 1995 (a period not covered in the graph).

From then until last year the boom attracted people into the workforce who weren't in it before. Thus, employment increased without the unemployment rate declining as much as previous relationships would have indicated.

(Remember that in 1996 Prof. Krugman dismissed the notion that 4% growth for five years was a credible possibility, because those relationships indicated so much growth would reduce unemployment to only 1.5% -- which clearly was not plausible. See his essay "The Four Percenters", in _Accidental Theorist_, which used this argument to rather rather beat up upon those who believed in 4% growth.)

BLS data for 1994 gives 66% of the population as "in" the labor force and 33% as not in it. From then until 2001, it gives 77% of the increase in population as being in the labor force and only 23% not. The increase in the size of the labor force relative to the population -- which provided jobs to persons who had not been counted among the unemployed previously -- would have tended to allow the increase in employment that we enjoyed during our actual five years of average 4% growth, without any decline of unemployment to the likes of 1.5%.

In 2002, however, the process sharply reversed. The BLS gives the year's marginal increase in the population as being only 30% in the labor force and 70% not. (The total labor force participation rate suddenly dropped by almost 0.5%.) Of course, those who aren't in the labor force aren't counted in the unemployment numbers. So that would allow a slowing of employment growth without the expected corresponding increase in measured unemployment.

IOW, it may be a mistake to assume a decade-long 1.5% average growth rate for the labor force is a *steady* growth rate, and calculate expected unemployment rates from there.

The growth rate accelerated around 1995 and caused Krugman to underestimate the unemployment rate that would accompany an accelerating economy. Now the growth rate has sharply slowed and caused the reverse effect -- an overestimate of measured unemployment accompanying a slowed economy.

Maybe. I haven't crunched the numbers in detail to confirm that this covers the whole thing. But it looks like a logical candidate as an explanation of the statistics.

Posted by: Jim Glass on November 2, 2002 10:55 PM

Another view, Barron's on the employment numbers:
~~
The flat trend in the unemployment rate continues to surprise many analysts. When joblessness fell to 5.7% in August, a New York Times headline called it "meaningless," citing informed sources who predicted it would soon vault above 6%. Now that it's still at 5.7%, analysts just call it perplexing.

But it shouldn't be. For one thing, GDP growth has averaged an annual rate of 3% over the past four quarters, and not so long ago that rate of growth would have been fast enough to push the unemployment rate lower. Only in this era of higher productivity does 3% growth mean that joblessness holds about steady.

The only reason to question the validity of the unemployment rate is that it doesn't square with the anemic uptrend in payroll employment tracked by the Establishment Survey of the jobs report. But the jobless rate, which comes from the Household Survey, may raise more questions about the payroll data than the other way around.

For one thing, payroll employment can be subject to huge revisions, far more so than household employment. And when the economy is in a growth phase, these revisions are more likely to be up rather than down. So it may be that the apparent weakness in payroll employment will eventually be revised away. Unfortunately, we won't know that for sure until June 2004, when the payroll data from April 2002 through March 2003 finally gets benchmarked to the "universal count" of the ES-202 data.

www.barrons.com

Posted by: Jim Glass on November 3, 2002 08:38 AM

Paul Krugman did not allow for the rise in productivity growth after 1995. Growth could and did accelerate and pulled down unemployment to numbers that would have been taken as quickly spurring inflation by economists before 1995.

Perhaps many who entered the labor force after 1995, did so simply because the opportunity was there and now are out of the labor force and not actively looking for jobs.

There simply does not seem to be the concern with loss of employment that was evident in 1981 or 1991. The economy is weakening, there have been huge investment losses, but the unease is not what I would expect.

Curious period.

Posted by: on November 3, 2002 09:17 AM

>> Perhaps many who entered the labor force after 1995, did so simply because the opportunity was there and now are out of the labor force and not actively looking for jobs. <<

Yes, that's exactly the point. During the last years of the boom, as the employed percentage of the population hit an all-time high, the business press was full of stories about firms striving to recruit retirees and others who previously weren't in the market looking for jobs at all.

Employers were offering flex-time as never before, letting people bring their pets to work, and so on. Since those hires didn't come off the unemployment lines, their hiring didn't reduce measured unemployment by the amount that the ratios Krugman relied upon predicted.

If so, and the process has now reversed, these people who were not hungry for work to begin would be the first ones to leave work and the labor force too, reducing employment without increasing unemployment.

The way I count it, the half-point drop in the labor-force participation rate accounts for about one million people leaving work without being counted among the unemployed, and that accounts for the difference between today's 5.7% unemployment rate and about a 6.2% rate -- covering most of the gap up to Prof. Delong's expected 6.5% rate.

>> There simply does not seem to be the concern with loss of employment that was evident in 1981 or 1991 <<

Well, unemployment today isn't nearly as high as it was in those cases. And to the extent that the people leaving work today are like those described above, there's not nearly as much hardship imposed on them as on the typical laid-off worker in the earlier recessions.

Posted by: Jim Glass on November 3, 2002 12:12 PM

>>> The way I count it ... <<<

Rather than just eyeball the data let's use a calculator.

Using BLS July data, the labor force participation rate has dropped 0.47% from its recent high, almost all of the fall coming in 2002. This indicates there are 1.01 million fewer people in the labor force in 2002 compared to if the rate hadn't fallen, without any corresponding increase in measured unemployment.

If we add these 1.01 million to the reported number of unemployed, and also to the total reported number in the labor force, the reported unemployment rate for July rises from 5.86% to 6.52%. Bingo.

Posted by: Jim Glass on November 3, 2002 12:45 PM

The big question is what precisely is behind the changes in the labor force participation rates.

Posted by: Atrios on November 3, 2002 07:18 PM

Barn-burning growth in farm jobs, would you believe?

**The household survey found that civilian employment fell by 271k in October after rising by 711k in September, with the unemployment rate edging up to 5.7% from 5.6% in September. Fumbling around on the BLS site, I stumbled across a breakdown between agricultural and non-agricultural civilian employment (it's always been there, I'd just never looked into it before). I was surprised to learn that "farm" employment (excluding forestry and fishing) is estimated to have increased by 227k (a lazy 7%) in October after a 110k (3.5%) jump in September, with non-farm civilian employment down by 498k in October after a 601k jump in September.

**It turns out that farm employment is estimated to have grown by 415k in the four months since June; from a starting point of 3.1m farmers, that's 13%, a 46% annualised growth rate! The tiny farm sector (in June it accounted for only 2.3% of all jobs) supposedly accounts for almost fully half of the 861k new civilian jobs generated since June. Put another way, the massive surge in farm jobs may have been an important factor depressing the published unemployment rate at a time of little or no growth in non-farm payroll employment.

**Of course, the rapid growth in farm jobs - the fastest in more than 50 years of data - seems implausible to say the least. I doubt the increase in farm jobs in the past four months comes within a bull's roar of 13%. Is anything real going on here? Or it it best just to walk away with the conclusion that the household survey data - including the published unemployment rate - are too erratic to be taken seriously? Any thoughts?

Posted by: on November 3, 2002 08:32 PM

>>One last thing: are there some demographic changes at work here? People born in 1945 are turning 58 years old. Isn't that the earliest retirement age for entitlement to Social Security in the US? When did the baby boom really start?<<

Nice one Jean-Philippe. I think this is getting to the heart of the problem. Just what demographic impacts are at work here. I don't know, since I don't know the US demographic very well. But some ageing factor or other is at work.

It would be worth checking to see just when Japan started this move (sometime in the early 90's I reckon) and then we can start moving to a more detailed model of the deflation mechanisms. Ater all add productivity (capital deepening) make the output gap calculation, and there you have it, structural deflation.

>>Yes, that's exactly the point. During the last years of the boom, as the employed percentage of the population hit an all-time high, the business press was full of stories about firms striving to recruit retirees and others who previously weren't in the market looking for jobs at all.<<

This post of Jim Glass's is also very much to the point for the non-demographic component.

Posted by: Edward Hugh on November 4, 2002 02:53 AM

Have to say Jim Glass is persuasive. Data do that for a person. I take an occasional look at the monthly immigration statistics. Sure enough, they show a slowdown in immigration applications and approvals compared to a year ago. If that is also reflected in illegal immigration (why leave home for no job?), then there are fewer potential employees showing up (not all reflected in the data, of course). The same influences would work on households. If the blandishments of employers were once considerable and now are not, some folk might choose to stay home. I am also hearing dire stories from unemployed friends. Interviewers tell them that a year or two ago, they were forced to take on people they found outright objectionable, people they had no hope would ever live up to standards. Now, some applicants are superstars for whom there was no longer any work. With competition like that, some people really have no hope of landing a job, so why be an active seeker?

I have to wonder though, why that would lead to a higher job count from households than from establishments over an extended period? Over the past six months, the household survey shows an average 156k monthly rise in employment, while the establishments survey shows just 36k per month. The good professor says statistical discrepancy doesn’t explain the different impressions left by the household and establishments data, but I’d like to know how he reaches that conclusion. Brad?

I probably have probably mentioned the problems with small business and self-employed data in the establishments survey somewhere on this website. I know I have somewhere. The adjustment in the estimate for smaller establishments, the “plug” I think it’s called, to reflect changed economic conditions, works with a lag. If the economy stopped contracting late last year, and if the demands on those currently employed have become insupportable, then the plug, based on last year’s conditions, may be missing some employment gains. That notion does not conflict with what Glass has to say. I like a simple, single explanation, but sometimes more than one thing is at work.

Posted by: KHARRIS on November 4, 2002 05:45 AM

Dear Brad

A very interesting puzzle. I have a guess (doesn't fit last year too well). Late 90s had very unusual employment growth. Even boarderline in the labor force people began working. This gives some slack (well some spare taughtness in standard labor market jargon). Thus lot's of easily discouraged workers have recently left the labor force. The story is same amount of discouraging experience (duration of unemployment say) has a bigger effect on lf participation if it is very high.

Posted by: Robert on November 10, 2002 02:32 PM

Dear Sir/ Madam

I am Pro-active " can do attitude",I do have exceptional people skills,high energy levels and a bright and helpful pleasant personality with positive work attitude.
i am able to work independently and under pressure.
yes i am have strong interpersonal and communication skills and Fluent in English .
and I possess valid driving license.

yes I am looking to be rewarded and given the opportunity for personal and professional growth with ongoing support and training.

I request you to kindly give me suitable opportunity to work to the best of my esteemed organization.

Thanking you
regards
Pathan Parvez

Encl cv

Curriculum Vitae

PERSONAL DETAILS:

Name :- PATHAN PARVEZ MUBINKHAN.

Father’s Name :- MUBINKHAN I. PATHAN.

Permanent Address :- 42, Vohra Colony,
Ajwa Road,
Vadodara. => 390 019.
(Resi) Phone No => 0265 - 560434.

Passport Information :- Passport No :- A 7514189.

Date of Birth :- 30th March 1975.

Sex :- Male.

Marital status :- Unmarried.

Nationality :- Indian.

Languages known :- English, Hindi, Gujarati.

Driving Licence No :-GJ / 6 / VD / 1993 / 70500 . (L.M.V)

Email :- Parvez42@hotmail.com

OBJECTIVE :-

To seek challenging avenue where my experience matches the
organization’s growth and strength and to contribute for the chievement of organizational goals.

educational Qualification :-
# Passed B.Com From M.S.University, Vadodara, in the year 1996-97.
# Passed M.Com From M.S.University, Vadodara, in the year 1999-2000.

Additional Qualification :-

# Cleared successfully Microsoft Certified professional - Solution
Developer, from Microsoft Corporation - California,U.S.A. in year 2000.

(Subjects and Marks)
[1] SQL Server 7.0 (70-029) Designing and Implementing Database Scored =>775/1000.
[2] Visual Basic 6.0(70-175) Designing and Implementing Distributed
Applications. Scored => 928/1000.
[3] Visual Basic 6.0(70-176) Designing and Implementing Desktop
Applications. Scored => 928/1000.
[4]Solution Architecture(70-100) Analyzing Requirements. Scored => 820/1000.

# Diploma in R.D.B.M.S :
• Detailed Study on client server Technology
• Structure query language as back end tool
• Procedural language /Structure quary language
• Developer 2000 as front end tool
• Forms, Reports
• Application Development through Oracle Server.

# Diploma in Office Automation :
• Computer Fundamentals
• FoxPro Fundamental
• Local Area Network
• Basic Hardware

# Web Designing :

• HTML(Hyper text Markup language).
• DHTML(Dynamic Hyper text Markup language).
• Java Scripting.
• Adobe Illustrator
• Adobe Photoshop
• Adobe Premiere(video Editing)
• Gif Animation
• Macromedia Fireworks
• Macromedia Flash(2-D Animation)
• Macromedia Dream weaver
• Sound Forge(sound Editing)

Currient Activities: Presently I am doing Post Graduate Diploma in
Journalism and mass Communication From IGNOU.

Experience:
presently working with Super marketing as a Business Development Executive at Baroda,India. Since last 5 year.

All the information given above is true to the best of my knowledge and
brief.

Thanking You,
Yours faithfully,

PARVEZ M. PATHAN

Posted by: parvez pathan on November 30, 2002 10:06 PM

Name:


Email Address:


URL:


Comments:


Remember info?