Job Market Glass Looks Half Empty

Either way, economists see little likelihood for a jaw-dropper.
Publish date:

To generate a reaction in the polls or at the

Federal Reserve

, Friday's employment report for September needs to show a labor market that is either falling apart or exploding.

And while most economists say that's not likely to happen, recent labor market indicators suggest the risks may lie on the side of a weaker number. What's more, White House hopes for a sweeping, upward revision of previous months' payroll data might also be dashed, some economists say.

According to consensus estimates, nonfarm payrolls rose by 150,000 last month, up slightly from 144,000 in August. The unemployment rate is expected to have held steady at 5.4%, and average hourly earnings are slated to rise 0.3%.

"If it is not startling outlier on either side ... Friday's numbers will not have much effect on the outcome in November," said Vincent Boberski, director of fixed-income research and a senior economist at RBC Dain Rauscher.

The jobs report is the last before the Nov. 2 election, and several polls show a statistical dead heat between President Bush and Sen. John Kerry. Meanwhile, futures markets are pricing in another 25-basis-point interest rate hike at the Fed's policy meeting on Nov. 10. For opinions to change about either issue, analysts say a surprising result from the jobs data will be required.

A payroll number of around 250,000, as projected by Wachovia Securities economist John Silvia, would likely cement expectations for a fourth rate hike this year, raise the odds for a fifth and perhaps allow Bush to regain some of the lead he lost to Kerry after the first debate.

In contrast, a negative number, as projected by Argus Research economist Richard Yamarone, could throw another rate hike into question and add to Kerry's recent momentum.

"Neither, however, looks especially likely," said Boberski. "The model that we use pegs a slightly below-trend number of 115,000 for September. The unemployment rate should remain at 5.4%. Those are probably not numbers that will stick in anyone's heads."

Recent employment statistics offer mixed clues about the state of the labor market last month. The Institute for Supply Management reported that employment in manufacturing and in the services industry both increased in September.

But some analysts say these measures have not been accurate predictors of job growth lately, and they note that the number of people who found jobs harder to get last month jumped to the highest level since April, according to data from the Conference Board. The percentage of people who saw jobs as plentiful declined.

In addition, planned job cuts soared to an eight-month high in September, according to outplacement firm Challenger Gray & Christmas. Layoffs, which are not seasonally adjusted, jumped 41% last month compared to the same period last year and were up 45% from August.

Among those announcing job cuts recently are




Computer Associates

(CA) - Get Report



(CAL) - Get Report



(PEP) - Get Report





More evidence of a weak job market can be found in an article by

Market News International

. The "Reality Check" story, which has a good record predicting the direction of nonfarm payrolls relative to consensus estimates over the past seven years, said recruiters reported a soft or flat job market in September and that employers have adopted a "wait and see" attitude.

What's more, weekly unemployment claims have been trending higher in recent weeks, although analysts say the data have been skewed by hurricanes.

Ethan Harris, senior economist at Lehman Brothers, is calling for job gains of 175,000 in September, because he says bad weather didn't depress job growth as much as other economists seem to think.

Hurricane Frances came before the employment survey was taken and Hurricane Ivan arrived on the Friday of the survey week, he said.

"Anyone employed before Friday counts in the survey, thus Hurricane Ivan would affect net job creation only in a small part of the country for one out of the 30 days between surveys," he said.

Still, more pessimistic economists note that September has been a poor month for payrolls over the past five years, and that businesses were particularly cautious last month.

"More and more employers are complaining about the inability to take on new workers, given the skyrocketing cost of heath insurance and benefits," said Yamarone, who thinks jobs shrank by 10,000 in September.

Record high energy prices also have added to a sense of caution, analysts say. William Poole, president of the Federal Reserve Bank of St. Louis, said high oil prices "have probably taken a bite out of the recovery" and that there is "slack" in the labor market.

In addition, Dallas Fed President Robert McTeer said there is "lots going on in the world that creates uncertainty and causes business people to be cautious about hiring and cautious about new investments."

Whatever the jobs data reveal Friday, the report should help to settle one of the arguments raging all year about whether the establishment survey or the household survey has been a more accurate gauge of job creation.

Along with the payrolls for September, the government will release preliminary estimates of benchmark revisions to payroll employment for the period of March 2003 to March 2004.

Each month, the Bureau of Labor Statistics estimates employment figures from a survey of 400,000 businesses, but the projections aren't always accurate. So, once a year the government corrects these estimates on the basis of more complete data from unemployment insurance tax records. Estimates of the benchmark revisions are announced in October, and the actual revisions come in February.

President Bush's Council of Economic Advisers has predicted that payroll employment could be revised upward by 288,000 jobs in the 12 months ended March, and perhaps as much as 384,000, according to

The Wall Street Journal

. But other economists say those figures are optimistic.

J.P. Morgan economist Bill Sharp called the assumptions "aggressive" and said he is looking for an upward revision of 150,000 to 200,000 in total, or about 15,000 per month.

And Lehman Brothers' Harris said payrolls could even be revised down during the period, just as they were over the past few years.

"For three years in a row, a number of analysts have argued that the payroll numbers were too low," and that the household survey better reflects what is going on in the job market, Harris said. "However, this argument has been strongly rejected by the hard data, and payrolls have been revised down rather than up."

According to the household survey, the unemployment rate has fallen to 5.4% from a high of 6.3% last June. But Harris said these numbers are flawed because they assume too high a growth rate for the population.

The household data are derived from calls to around 60,000 homes. The government then extrapolates estimates of job growth and the unemployment rate for the entire population. In contrast, the payroll data are based on a survey on 400,000 businesses.

Unemployment insurance records "show slightly less job growth than the published payroll numbers" for the year ending in March, Harris said, adding that "new data from the IRS show lower wage income, suggesting either fewer jobs or lower pay per job."