"Economic Growth Remained Strong in the Fourth Quarter," is the headline over the New York Times story. A rather strange headline for a quarter during which output grew at a slower pace than expected and employment did not grow fast enough to soak up expected growth in the labor force.
"Though government data on payrolls showed almost no job growth in December, government surveys of households have suggested that hundreds of thousands of jobs have been created," reads a passage near the story's end. While the payroll survey did show employment growth of only 1,000 the household survey did not show "hundreds of thousands" of jobs created during December--it showed that 50,000 were lost. Over longer time periods the household survey does indeed paint a much brighter picture than the payroll survey--but not during December.
"The contradictory evidence remains a puzzle, but one theory is that many people who no longer work at companies are now self-employed as consultants and service providers who work out of their own homes," reads the sentence immediately following. But the divergence since January 2001 between the household survey and the payroll survey has been 2.6 million, and the household survey itself reports that self-employment has grown by less than 15% of that divergence.
The New York Times is simply confused, and inaccurate.
Perhaps the first thing the New York Times should do is settle on a definition of a "strong" economy. I would argue that a "strong" economy has to have a relatively high (or at least a rapidly rising) level of employment relative to the trend growth of the labor force. And the second thing it should do is bring itself to write the one sentence--saying that productivity growth in the fourth quarter was strong, employment growth was weak, and output growth was pretty good but still worse than expected--that would give its readers the right clues to what is going on:Economy Remained Strong in 4th Quarter, U.S. Says: The nation's economy grew at an annual rate of 4 percent in the final quarter of last year, much slower than the torrid pace of last summer though still rapid by historical standards.... Though the growth remained strong, it was weaker than many economists had been predicting. The pace also raised doubts about whether job creation will increase rapidly enough to reduce the numbers of unemployed significantly by the time President Bush faces re-election in November.... [M]any forecasters had been predicting that growth would hit an annual rate of 5 percent in the fourth quarter and remain well above 4 percent for the rest of this year....
[Analysts] they were much less certain about the outlook on unemployment, with some predicting that job growth will remain modest this year and fall well short of replacing the 2.5 million jobs that disappeared during the last three years. "We are still growing about twice as fast as Japan and almost three times as fast as Europe," said Nariman Behravesh, chief economist at DRI/Global Insight, a forecasting company in Lexington, Mass. "Nevertheless, 4 percent is probably not strong enough to have a huge impact on the jobs market. We are likely to see some gains in employment growth, but they will be modest."...
The biggest restraint on job creation has been an almost unprecedented surge in productivity, the amount of output that companies can squeeze from the same number of workers. Productivity has risen at an annual rate of about 5 percent for the last two years, and shot up 9.2 percent in the third quarter of last year.
If productivity continues to climb at even 5 percent, something that most economists consider improbable, the economy would have to grow by more than 4 percent a year to generate a drop in unemployment. The jobless rate was 5.7 percent in December. White House economists, fixated on the issue of job creation for both political and economic reasons, contend that the picture is already better than it appears.
Though government data on payrolls showed almost no job growth in December, government surveys of households have suggested that hundreds of thousands of jobs have been created. The contradictory evidence remains a puzzle, but one theory is that many people who no longer work at companies are now self-employed as consultants and service providers who work out of their own homes.
And there are the things in the article that are just plain erroneous. For example:
If productivity grows at 5% per year, we need GDP growth of more than 6%--not 4%--per year to produce a falling unemployment rate.
The household survey has generally given a more optimistic picture of the labor market than the payroll survey--but not in December. In December the 1,000 increase in estimated payrolls was accompanied not by "hundreds of thousands" of new jobs in the household survey, but by a fall of 50,000 in household survey-concept employment. (And a one-month comparison of trends in either of these series is a silly exercise anyway.)
And why don't people talking about the theory that the big gap between the payroll survey and the household survey since 2000 is due to the rising number of the self-employed mention the household survey estimates of the self-employed? The household survey asks people if they are self-employed. And the household-survey estimate of the numbers of the self-employed has risen by only 330,000 (from 9.14 to 9.48 million) since January 2001--only a small fraction of the 2.6 million difference between the payroll survey trend and the household survey trend. Isn't it worth pointing out that the household survey itself lends no support to the theory that changes in self-employment are an important factor in understanding the current labor market?
Posted by DeLong at January 31, 2004 01:46 AM | TrackBack | | Other weblogs commenting on this post
It's important to remember that productivity is calculated using the payroll survey numbers, not the household survey. Rather than saying that high productivity demands higher job growth to reduce unemployment, I think it makes more sense to believe that the productivity numbers are overstated because the payroll survey is understating current job growth.
Posted by: Bruce Bartlett on January 31, 2004 05:35 AM
The GDP data implies that 4th Q productivity growth will be about 2% -- 4% real GDP
vs 2.2% gain in hours worked.
Non agricultural self employment is now equal to
7.0% of non-ag employment .
Since 1967 self employment has averaged 7.2%
of non-ag employment
In the 1980s & 1990s this ratio ranged from 7.5% to 8.0%.
On this basis there is nothing unusual about the current level of self employment-- actually by historic norms it is low.
Posted by: spencer on January 31, 2004 06:08 AMThis is clearly a situation where the rate of change is the story, not the absolute value. But the Times is too thick to see it.
Posted by: Bob H on January 31, 2004 07:26 AMJust curious:
When you meet someone at a party who says he's self employed, you think, "unemployed but not willing to admit it to anyone, including himself."
What strategies does the household survey use to guard against such prideful delusion?
Posted by: Yesh on January 31, 2004 07:47 AMIf the household survey is just asking people whether they're self-employed or working as an employee, that's not going to capture most of the shift to off-payroll work unless respondents are very clear on what self-employment is. Your average temp or permalancer certainly doesn't feel self-employed -- they work 8-6 like everyone else, get no benefits like almost everyone else, have no control over the conditions or content of their work...
The only visible difference between regular employment and self-employment for a lot of people is the form that comes in January. (Well, that and even more terminal uncertainty, should that be possible)
>>Rather than saying that high productivity demands higher job growth to reduce unemployment, I think it makes more sense to believe that the productivity numbers are overstated because the payroll survey is understating current job growth.<<
But the value-added by the self-employed *should* be excluded from non-farm business value-added...
Posted by: Brad DeLong on January 31, 2004 08:06 AMThe NYT muffed this story (as they do most other economics stories) for two reasons:
1. It's hard for a J-school grad to work with numbers, and even more difficult when those numbers involve a field with which the reporter has no real training or expertise. And
2. If the Times accurately reported in the state of the economy, it could result in real trouble for Bush's re-election. That goes against both the officially sanctioned propaganda line and the company's financial interests.
Posted by: Derelict on January 31, 2004 08:10 AMI am one of those that would answer self employed, because the $2,000 to $3,000 a year I scratch out of a catering business is all of the income I have. Fortunately my wife is a school teacher and I have been thrifty enough during my "employed life" to amass a nestegg. I am a college grad who has not "worked" for more than two years. My son graduated college in computer science (with huge debts). He has been living with my brother for 5 months in Seattle looking for work. THERE IS YOUR TRUE HOUSEHOLD SURVEY
Posted by: Don Beal on January 31, 2004 08:45 AMI assume that Bush's minions are faking that statistics, just as they fake or manipulate every other area of governmental information. I doubt that productivity is increasing in any way other than that employers are forcing employees to work more unpaid hours.
Posted by: Rich Puchalsky on January 31, 2004 08:59 AMIs Andrews a business writer or a political writer, or an essay writer? In the second paragraph of the article, he raises the impact of continued lack of job growth on the Bush campaign for reelection. A few graphs in, still on page one, he quotes Bush's empty comments. Later, he quotes Democrats about making tax cuts permanent.
He includes what looks to me like a straight pick-up from a White House press release, or economists talking point, clothed in the guise of his own thought, about the difference between the two employment surveys. He closes with an ambiguous anecdote about Brunswick boat dealers inventories, choosing the positive side.
No wonder no one knows anything. Any one of these lines of thinking would make an article.
Posted by: masaccio on January 31, 2004 09:00 AMWe have a most serious labor market problem and have wasted billions and billions of fiscal policy dollars with no real regard to job creation and wage and benefit improvement. When I hear analysts talk of the lovely move to self-employment that is masking an improving labor market, I am deeply annoyed.
Where are the secure, well paying, fair benefit jobs? Economic growth is fine, corporate profits are fine, upper management compensation is absurd, and middle and lower wage labor is being pressed and pressed. The lower the wage category, the more general pressure. But, there is a busy industry of analysts to tell us all is well. Yuch.
Posted by: anne on January 31, 2004 10:17 AMThe CBO predictions in the latest report assume an average growth rate in 2006 and beyond of about 2.7 percent. Has anyone reworked the prediction numbers for higher growth rates? Like 3 or 4 percent? Would such numbers be unreasonable?
Posted by: tbrosz on January 31, 2004 12:32 PMI've got a slightly off-topic question that's been puzzling me.
Will dropping the 13-week emergency unemployment benefits extension result in a decrease in the published unemployment rate?
Posted by: Ellen1910 on January 31, 2004 01:49 PMhttp://www.cbpp.org/1-29-04ui.htm
Unmet Need Hits Record Level For the Unemployed - 1/29/04
Due to the end of the temporary federal unemployment benefits program, the estimated number of unemployed running out of regular unemployment benefits and qualifying for no further aid will be higher in January than in any other month on record, and is also expected to hit a record level for the first half of the year.
Posted by: anne on January 31, 2004 02:10 PMhttp://www.cbpp.org/1-29-04ui.htm
An estimated 375,000 unemployed individuals are exhausting their regular unemployment benefits in January without qualifying for any further assistance — and are receiving neither a paycheck nor unemployment benefits. Based on the latest data, nearly two million unemployed workers are expected to be in this situation during the first six months of 2004.
In no other month on record — and in no other six-month period for which data are available — have so many unemployed workers exhausted their regular unemployment benefits without being able to receive additional aid. This finding holds even if the number of exhaustees in previous years is adjusted upward to reflect the growth in the labor force since then.
Ellen - I do not know how this may effect the "rate" of unemployment.
Posted by: anne on January 31, 2004 02:13 PMEllen, yes it will. Only those collecting unemployment are counted as unemployed.
Posted by: Chuck Nolan on January 31, 2004 04:05 PMChuck I'm not sure you're correct; I think non-discouraged workers who are still looking for work are counted as unemployed even if their unemployment benefits have been exhausted. But something like that is the gist of my question.
Isn't it possible that pushing long-term unemployed workers off the rolls will increase the numbers of discouraged workers who are not counted as unemployed.
Such an imaginative way of reducing the unemployment rate in an election year, isn't it? But of course, we won't hear any of this from the Times.
Posted by: Ellen1910 on January 31, 2004 05:43 PMEllen, I think I am right. I heard this from an Econ professor in a PhD level course.
How else would they count unemployed? If they are not receiving benefits, where's the data coming from? Where else and how else are they being counted?
Posted by: Chuck Nolan on January 31, 2004 09:18 PM"Each month, 1,500 highly trained and experienced Census Bureau employees interview persons in the 60,000 sample households for information on the labor force activities (jobholding and jobseeking) or non-labor force status of the members of these households during the week that includes the 12th of the month (the reference week)."
http://stats.bls.gov/cps/cps_htgm.htm
Posted by: Ellen1910 on January 31, 2004 10:58 PMThere's nothing to gain and nothing to lose.
Posted by: Dulabaum Nina on May 2, 2004 03:41 PMThe function of the artist is to provide what life does not.
Posted by: Ahmad Saif on May 3, 2004 02:51 AM