April 02, 2004

Finally, Some Good Labor Market News!

Finally, some good labor market news!

WSJ.com - U.S. Payrolls Grew in March At Fastest Pace in Four Years: Job creation jumped to a four-year high in March as the economic recovery finally filtered through to the long-suffering jobs market. The Labor Department said Friday that nonfarm payrolls soared by 308,000 in March from February, the biggest monthly gain since April 2000, at the height of the stock-market bubble. It also revised upward the tally for January and February, showing total job creation in those months of 205,000 instead of the previously reported 118,000. The unemployment rate, which is derived from a different survey than payrolls, rose a tenth of a percentage point to 5.7% as 179,000 people re-entered the labor force in search of jobs...

Mr. Bush's chief economist, Gregory Mankiw, said the data confirm implications from other recent statistics that "the economy is on track for a robust recovery." He added that the recovery is a "result of pro-growth fiscal policies" and low interest rates. "It is a textbook recipe for how you get out of recession."

Greg was, however, careful not to say that even the March pace of employment growth was less than the 320,000 per month that he forecast last February would be the average monthly gain between October 2003 and December 2004...

Posted by DeLong at April 2, 2004 08:06 PM | TrackBack | | Other weblogs commenting on this post

Is this good news for the Bushies, and bad news for the Kerry camp? Well, not if you look at the numbers in some depth.

Take a look at this analysis of the March employment data at:


Posted by: rickfman on April 2, 2004 08:18 PM


Pardon my complete lack of knowledge here, but would someone please tell me how the number of 179,000 (people re-entering the job-seeking market) is derived? I understand that they re-enter...how is it ascertained by the Gov't?


Posted by: john on April 2, 2004 08:41 PM


It's just survey sampling. The BLS has a good primer here:


and here:


As for this specific news - you can drop a fair chunk from the numbers for the California workers coming back and there is a one time bonus of construction workers going back after lousy weather in February. The interesting thing, in my mind is the difference between the two surveys. When you look at them unadjusted both are moving in the same direction but digging deeper you find that self employment is down (which is one reason why the pop survey is lower) and that payroll employment is up. I actually see this as good news. Through most of the 90's the gap betweent he payroll and household surveys decreased and I believe that a lot of people employed as self employed are not well employed - so getting on a payroll is good.

The figures aren't quite as good as the headline number but overall it's a good report and the movement to payroll employment is very positive imo.

Posted by: Ian Welsh on April 2, 2004 09:39 PM


What I can't understand is how, given that it is widely accepted that both low interest rates and deficit spending are stimulative, anyone can point to the upturn in the economy as proof that Bush's tax cuts "worked". Given historically extremely low interest rates and the biggest deficits ever, what's amazing is that it took so long to see improvement. Bush should get no credit for something that should have happened long ago, if his tax reductions had been properly structured.

Posted by: jymm on April 3, 2004 04:28 AM


Dow Jones Newswire's Dr. John McAuley admonishes investors that "March Jobs Gain Welcome, But Hold The Bubbly"

McAuley, who holds a Ph.D. in international trade and macroeconomics, and has been teaching Economics at Fordham University since 1974, noted "the headline increase was nearly the only upbeat statistic in the entire jobs report."

Beyond the headline number, he observed:
-The unemployment rate jumped 0.1 percentage point to 5.7% as 182,000 people joined the jobless ranks, eclipsing the 179,000 increase in the labor force. There were 284,000 workers who lost jobs in March, a decline of 49,000 first-time job seekers, and an unchanged number of workers returning to the work force.

-The workweek shortened by six minutes to 33.7 hours in March - an indication that the existing workforce was used less intensively.

-Indeed, when the number of production workers are multiplied by the hours they worked to get an indication of total labor input, the resulting index of aggregate hours worked actually declined by 0.1% last month.

-Average hourly earnings only edged up by 0.1%, so, combined with the decline in total hours worked average weekly earnings fell by 0.2% in March, suggesting that there may have been more jobs, but less income.

-In percentage terms the jobs gain in March amounted to 0.2% compared to the expectation of an increase of 120,000, or 0.1%. If the payroll data were regarded like any other economic indicator - which measure growth or contraction by percentage change - this increase would be interpreted as roughly in line with expectations.

-There was actually a small (1,800) decline in temporary help service workers, a category that is often looked to as a leading indicator of wider hiring gains, particularly in manufacturing.

-In addition, the factory workweek - one of the components of the index of leading indicators - fell by six minutes, so aggregate hours worked in manufacturing declined by 0.3% in March, an indicator many economists use for forecasting industrial production.

Improving, but still somewhat mixed . . .

Posted by: Barry Ritholtz on April 3, 2004 05:01 AM


Max Sawicky has a good blog post on this:

Better yet is this Galbraith ("the younger") article at Salon (registration required for full story):


Galbraith has a graph which, if the numbers behind it are valid, shows that those absurdly optimistic employment projections Bush et al. keep making (and which Brad and Paul Krugman decry) *are merely a return to trend*.

You can view the Galbraith column even if you don't have a sub to *Salon* (I myself don't have one) by "watching a short ad," which for me was rather painless (i.e., took almost no time, but I do have DSL).

Posted by: liberal on April 3, 2004 05:59 AM



I could not agree more. As an academic and educator, I find the media and public reaction to this kind of news incredibly depressing; it demonstrates the total lack of ability of our journalists -- and many pundits -- to think rigorously about causality. ;p

Posted by: dn on April 3, 2004 06:16 AM


Can some of you economist types assure us they are not "messing with the raw data". This adminstration lies with impunity and it just would not surprise me to that Mrs. Mitch McConnell cooked the books for George et. al.

Posted by: Don Beal on April 3, 2004 07:49 AM


From John Kerry's radio address:

"There is not a single month of this administration that has seen the creation of a single manufacturing job."


Democrats and paleocons (like Pat Buchanan) seem to think that manufacturing jobs are particularly important. Is there a good economic justification for this position?

Posted by: Michael on April 3, 2004 09:10 AM


Michael, 'could be that the manufacturing jobs pay salaries/wages from which taxes are deducted, unlike the self-employment which has no such certainty.

Posted by: calmo on April 3, 2004 09:29 AM


I actually regard the unemployment rate increasing as good news - more people looking for work. Of all the job indicators the unemployment rate is one of the most worthless - it can decrease when things are getting worse and often increases when things are getting better - but it's the headline number that people report the most.

The decrease in time worked (except in the Manufacturing sector) doesn't strike me as necessarily a bad thing - it could just indicate that the amount of overtime being asked for has been reduced because of hiring.

It's not an outstanding report, but it's a good one. Tune in for the next few months to see whether employment really improves (and with it Bush's chances for reelection.) A lot of that probably still depends on the housing market - which I believe is in a bubble in many parts of the country (I know others disagree) - if Greenspan can keep it inflated till the election, great. If not, Bush can kiss his reelection chances goodbye.

Posted by: Ian Welsh on April 3, 2004 10:39 AM


Oh - on massaging by the BLS. I had the same suspicion, so I created my own very crude seasonal adjusted index (just a 10 year population weighted moving average of year on year changes) - which wound up generally showing more people employed than the BLS does but was, in fact, fairly close to their numbers. I have a lot more sympathy for the BLS after that little exercise.

The real manipulation takes place openly, most notably in the inflation indices with their quality adjustments, geometric weighting and reduction of the weighting of energy (well ok, that last one they did without telling anyone specifically what they were doing.) Of course, many economists agree with the changes to the inflation measurements - it's not clear that it was politically motivated and if it was it was motivated by a mid 90's desire to reduce the payments that are indexed to the inflation indices, especially the CPI.

Posted by: Ian Welsh on April 3, 2004 10:48 AM


I'm not intimately familiar with the data, but if this is actually good news, (economically) and I'm inclined to believe that it is, I'm really just surprised that it took so long.

From a political standpoint, I'm not suprised, but nonetheless more than a little disappointed.

The Bush administration has employed or enjoyed every sort of fleeting economic stimulus one can imagine short of a monetary policy in the vein of Arthur Burns. They've had an unsustainable fiscal policy, a war effort, and a plummeting dollar. Even the tax cuts, having not yet been paid for, will only boost the economy in the short-run, if not by much and not cost-effectively.

If such aren't sufficient to instigate a recovery, they would certainly inflate any recovery occuring autonomously as a simple movement of the business cycle. I certainly don't want to make any predictions, but I'd hardly be surprised to see an upswing between now and November entirely independent of policy or politics.

So as much as I'd like to see Bush lose, I'm not counting on the ostensible state of the economy to do it for Kerry. The administration has demonstrated a willingness to lay down every cheap policy trick it can to buy economic indicators for the short-term, and I don't see a particular reason to believe that the business cycle will continue to work against them.

Posted by: Stu on April 3, 2004 10:52 AM



the emphasis on manufacturing jobs comes from a few different angles. For the paleocon contingent, manufacturing is a part of their nostalgia for an America in which everyone made steel and cars with the sweat of their brow and the branw of their broad American shoulders, and no little brown and yellow people in foreign lands stole ;our' jobs. For the Democrats, there's also a tinge of nostalgia - the blue-collar, white-ethnic (read Catholic), labor union vote was their traditional base, and they're stuck in the loop of appealing to them by highlighting their lost jobs.

In pure economic terms there are a few reasons why manufacturing jobs are important - manufactured goods are more readily exportable, thus helping our balance of trade, and they are considered more stable than retail or other service jobs becuase they typically require capital expenditure to create - business-owners need to expand their plants if they want to build more refrigerators, so that indicates a stronger belief in the long term economy and a less-transient workforce.

From my presepctive the real question is not service/ retail vs. manufacturing, it's permanent full-time vs. temporary part time, and in that perspective the new numbers don't look so great - more workers areaccepting part-time work becuase they have no choice (as opposed to making a lifestyle chjoice to work part-time) and temp hiring is on a long-term uptrend.

As I said in my post:


this is a complex set of numbers that require some analysis to interpret.


Posted by: rickfman on April 3, 2004 10:56 AM


Stu, I don't think the Bush administration has pulled any of the right levers in terms of jobs stimulation. There are a lot of levers that they pulled that do nothing at all for either the economy or jobs. Mostly, the Bush administration has overseen a massive transfer of wealth from the US treasury to wealthy Bush campaign contributors, financed by massive deficits. New treasury bonds have soaked up most of the "tax cut stimulus".

Posted by: bakho on April 3, 2004 12:11 PM


rickfman: focus on manufacturing

Manufacturing of goods and performing tangible services are the cornerstones of the economy (for a quite broad definition of "tangible"). The rest is for the most part an edifice erected on top of those for the purpose of managing, coordinating, brokering, etc. these basic economic activities; managing the managing activities, coordinating the management of the managing activities, etc.

Often times people forget that the original purpose of economic activity has been to faciliate and improve the conditions of human life and development, not to produce numbers on sheets of paper or a computer's disk.

Take out the manufacturing, and you end up with people taking in each other's laundry, and people managing the laundry-intaking. (Metaphorically speaking, and no offense to people performing services. I'm in that category myself.)

Posted by: cm on April 3, 2004 12:11 PM


OK, we have new jobs, and that seems to be good, but I wonder just how great these new jobs are:

"It is too early to celebrate," said Sung Won Sohn, chief economist at Wells Fargo in Minneapolis.

He noted that most of the increase was because of part-time workers. The number of people who worked part time for economic reasons rose to 4.7 million in March, up from 4.4 million the previous month.

"The average duration of unemployment has been lengthening, persuading unemployed workers to accept part-time jobs," Sohn said.

source: http://tinyurl.com/3cuh3

Posted by: DanM on April 3, 2004 01:06 PM


Historically, manufacturing jobs seemed important because they paid the workers better than retail jobs, or slinging hamburgers or doing nails. Nowadays, maybe that is less so, as the manufacturing jobs move off-shore. Since the capital to build plants is mobile, and goes farther in many other countries, there is little room to pay Amreican wages to locate here.

Posted by: masaccio on April 3, 2004 01:22 PM


Given the increase in both government and household debt during the Bush presidency, it'll be interesting to see what happens when (or if) interest rates return to non-deflationary levels. Even if a real jobs recovery materializes in the near future, we've still got a long climb out of the hole the Bushies have dug.

Posted by: Tom Marney on April 3, 2004 01:32 PM


Since we will probably continue to import, we need to ask what will we export? Those jobs won't be outsourced.

Right now I can think of only one area that is going to be almost totally resistant to outsourcing. Agriculture. As those Indians and Chinese get rich, they will want to eat more meat. I don't think they have the land to grow food for all those chickens, cows and pigs. Unfortunately, our agriculture does not need a lot of workers. Forestry fits in here as well.

We will also probably continue to export services.

Posted by: ____league on April 3, 2004 01:46 PM



I suppose that the Bush administration and its apologists will soon drop their insistence that Bush's tax cuts for the wealthy be made permanent. After all, a healthy economy is quite capable of generating more than enough capital through (gasp) robust profit margins to eliminate the need for subsidized capitalization through unsustainably massive government borrowing, as the economic history of the Clinton boom amply demonstrated. And the economy is really healthy now, right?

Posted by: Tom Marney on April 3, 2004 01:52 PM


___league: Don't forget the "export" of military services and equipment, which is a very strong segment of the US economy. It's even "delivered to your door" hassle-free, but we have a no-return policy. Perhaps I shouldn't be so facetious, but I couldn't resist.

But seriously, the case has been made that military protection is part of the non-financial compensation that the Taiwanese, Japanese, and Europeans receive from the US in exchange for their goods. There is probably not an explicit deal putting a dollar value on it, but it allows those countries to have a more light military & "defense" industry, and focus their resources on other things, like technology research and production of advanced goods. And I can only suspect that there is some tit-for-tat that results in favorable conditions for US businesses.

Note: I'm not saying it shouldn't be so; ultimately the lives of US soldiers and other population groups will be put at risk, and there has to be some payoff.

When it comes to international affairs, one contributing factor why (Central/Western) European societies can take the moral high ground is that the US will step up to the plate. It's a kind of moral hazard situation; I can refuse to pursue a military engagement in a conflict because I know that you will go for it. For example, if Western Europe would not have had the benefit of US nuclear "protection", they would have had to create a larger defense posture of their own (regardless of who started the arms race; it was not the Soviet Union that dropped the first nuclear bomb).

I may have gotten off-topic, but I think this general area has at least to be considered.

Posted by: cm on April 3, 2004 02:10 PM


Ian - good information especially the good news about the Southern Cal grocery store workers. Yep - those in LA are happy we are not completely reliant on long CostCo lines and overcrowded Trader Joes for food. But the Bush tax cuts can't be responsible for this increase in employment. After all - the strike started after the latest tax cut. Hmm? Maybe the tax cut caused the strike!

Posted by: Harold McClure on April 3, 2004 02:21 PM


Any bets on how much this figure will drop when the corrected figure comes out?

I remember a string of 6-8 months where employment rose, went back to zero with the correction, and rose the next month.

Posted by: Matthew Saroff on April 3, 2004 08:03 PM


Damn, it is painful to see my fellow liberals squirming so. You are so ticked at Bush saying Down is Up that you are now insisting that Up is Down. Shame on you. You are parallel-universe Karl Roves.

Face it. You picked the wrong fight and lost. Time to move on. Time to recognize that George Bush has little effect on the rate of employment growth in the USA. That is determined by the Fed, chance and the rhythms of the business cycle itself. You always knew that but you suppressed your reason in order to score political points. Shame on you.

And too bad. Your foolish posturing over jobs (and worse still trade) does raise the odds that the idiot will be re-elected. When a moderate sees John Kerry claiming, incredibly, that an Indian is about to steal American jobs, he will just laugh, shake his head, plug his nose and vote for the incumbent, who is stupid and crazy but not CERTAINLY wrong.

Try this instead. Use rational arguments against Bush that will stand the test of scrutiny. It is not like they are hard to come by. When BDl is not bean counting, he provides all the ammo you need.

Posted by: Gerard MacDonell on April 4, 2004 08:38 AM


Sorry Gerald, but that doesn't add up. Bush's economic policy has, in fact, been bad for jobs. And remember I'm one of the people who is arguing that this is a good job report. Bush provided a stimulus in the least effective fashion possible. This isn't rocket science - we know how to apply a fiscal stimulus to the economy in a way that gets the money into hands that will spend all or almost all of it. Bush didn't do that.

Also bear in mind, as Stirling has pointed out, that this is Bush's best job report - and is about equal to an average Clinton job report. It's not spectacular as some have said, it's just a good report.

Bush has mishandled the economy horribly and mishandled jobs handily and I will make this prediction - as measured by the establishment survey, by October, Bush's term will still show negative job growth. As measured by the population survey, by October, there will still be a lower labor force participation rate than there was when Bush took power.

Posted by: Ian Welsh on April 4, 2004 09:17 AM


Gerard MacDonell: While I have to agree that a number of people are knee-jerking, coming forward with accusations of number fudging when the numbers don't fit their expectations, etc. (and this leads me to the obvious concerns), it _is_ certainly true that the numbers are not painting a consistent picture, and people want to explain this to themselves.

For example, we have heard and some of us know anecdotes of how unemployed acquaintances have not been able to get their old pay level in a new job, or are employed below their level of expertise. "Available" signs are still being put up, companies are hiring very cautiously, etc. All of that causes doubt, not about the accuracy of the numbers, but what they really tell us.

Seasonal and other adjustments complicate things further; over a short range the numbers of the two surveys need not necessarily coincide. Here are some numbers from the report; difference Feb/Mar 2004, all numbers in thousands):

Gross household data: (NA not-adjusted, SA seasonally-adjusted), based on table A-1

Population Feb/Mar: +193 (no adjustment)
Labor force Feb/Mar: NA +371; SA: +179
Employment Feb/Mar: NA +307; SA -3
Unemployment Feb/Mar: NA +64; SA +182
Not in labor force Feb/Mar: NA -178; SA: +14

We see that the respective numbers add up nicely (labor force = employed+unemployed; population = labor force+not labor force). We also see large differences between NA and SA; for example, there is presumably always an employment peak in march, so the +307,000 is corrected to a slight decline.

On to table A-5, by class of worker and part time status:

Agri wage/salary: NA +58; SA -12
Agri self-employed: NA +8; SA -16
Non-Agri wage/salary: NA +473; SA +239
Non-Agri self-employed: NA -241; SA -288
Combined wage/salary: NA +531; SA +227
Combined self-employed: NA -233; SA -304

I did not list unpaid household/family worker as the SA data are not given, and the SA values in this table don't add up to the A-1 values, as they are independently adjusted.

Part time economic reasons: NA +104; SA +296
Part time non-economic reasons: NA -37; SA +106
Part time total: NA +67; SA +402

On to table A-8, unemployed by category:

Job losers/end of temp contract: NA +32; SA +284
Job leavers: NA +25; SA +9
Reentrants: NA +0; SA +0
New entrants: NA +6; SA -49

Reentrants are people who started looking for jobs (again and the first time, respectively). We note that the NA numbers add up to the NA unemployed; the SA numbers do not.

Now the establishment survey numbers (all preliminary):

Total nonfarm: NA +1,007; SA +308
Natural resources/mining: NA +7; SA +7
Construction: NA +160; SA +71
Manufacturing: NA +42; SA +0
Trade, transportation, utilities: NA +133; SA +73
Information: NA +3; SA -1
Financial activities: NA +18; SA +6
Professional and business services: NA +142; SA +42
Education and health services: NA +83; SA +39
Leisure and hospitality: NA +224; SA +28
Other services: NA +35; SA +12
Government: NA +159; SA +31

Analysis: The two surveys don't match up well; non-adjusted household employment is up 307,000, whereas payroll employment is up 1,007,000. Even when speculating that a good number of the previously self-employed were not in the payroll survey, and that there is a large number of multiple job holders, the gap still looks large.

What is striking is the relatively large drop in the self-employed; are those people who where doing contract gigs or looking for such, and were now hired?

The increase of "part-time for economic reasons" is approx. one-third of the increase in the newly employed. I have to look at historical series to see how this fits.

Any other comments on these numbers from people who are familiar with the matter?

Posted by: cm on April 4, 2004 01:05 PM


Prudent Bear has an article which cross-correlates some BLS numbers with other numbers: http://www.prudentbear.com/archive_comm_article.asp?category=Guest+Commentary&content_idx=31571
The article does not question the BLS numbers, but their interpretation in regard to other numbers.

Posted by: cm on April 4, 2004 07:24 PM


The ISM non-factory data just came in, and they support the impression of a strong showing from the economy in March. That fits well with both the payroll data and the earlier factory ISM results. I'm thinking that Q2 GDP may be a good bit stronger than the 4.5% or so median estimate. Tack strong ISM and jobs data on to the impact of tax refunds on April and May consumption and things look good.

For those who want to doubt the integrety of the BLS data, I would offer the following observations.

1) Lots of folks are involved in putting the jobs report together, and I am aware of no instance of a BLS statistician crying foul. Since, as civil servants, their jobs would seem at greater jeopardy from fudging the data than refusing to do so, there seem to be serious practical difficulties in fudging the data.

2) Factory jobs showing no growth, when the median estimate was for a modest rise, and when ISM data continue to show factory job gains, seems a pretty silly way to jigger the data. Kerry got right to work pointing out the absence of factory jobs.

3) A gain of 108k jobs is needed, in any given month, to register as a statistically significant employment gain. Given the magnitudes this suggests we should keep in mind when looking at establishments data, the unexpected lag in employment after several quarters of strong output growth, and the fact that this is just one month's as yet unrevised data, do we really need to look for jiggery-pokery to explain a 308k job rise?

Posted by: K Harris on April 5, 2004 07:56 AM



" As Alfred Marshall used to say, 'nature makes no leaps,' right? The numbers on one side of 2000, this graph shows, simply aren't comparable to the numbers on the other side of the divide. The value of this questionable upward shift in the labor force and employment numbers is approximately 1.3 million workers or jobs."

Posted by: Patrick R. Sullivan on April 5, 2004 09:14 AM


K Harris: I believe few in this forum doubt the integrity of the BLS data and processes; I certainly don't. My quibble is with the definitions and categories of what is measured, not only by the BLS, but also other econometric agencies. It is very tempting if category A is virtually impossible to measure (or even define properly) to measure the more easily accessible A' and claim that you can make significant assertions about A.

Applied to unemployment, people cannot even agree which category of people should rightfully be considered unemployed, so you measure one thing and publish the number. The number wil correctly mirror what you have measured, but extrapolating this to interpretations about widespread notions of unemployment becomes a stretch. And then while unemployment is a significant indicator, is it the most decisive one? Only if you make it so, in lieu of things that you can't or won't measure easily, like income distribution by worker category.

Posted by: cm on April 5, 2004 09:18 AM


I think Bush should not be president and I do not trust Cheney and Rumsfeld. I don't know about Kerry, but I think i will vote against bush. These people give me a really bad feeling

Posted by: online bingo mom on April 30, 2004 03:32 PM


I think Bush should not be president and I do not trust Cheney and Rumsfeld. I don't know about Kerry, but I think i will vote against bush. These people give me a really bad feeling

Posted by: online bingo mom on April 30, 2004 03:37 PM


Find the best online casinos, with information on their bonuses, payouts and popularity

Posted by: portal to online casinos at the online casino helper on May 30, 2004 08:37 AM


Online Casino Directory

Posted by: Online Casinos on June 23, 2004 03:00 AM


Post a comment