Confusing job surveys hint at moderate growth

• A guide to understanding Friday's report on job gains and the unemployment rate
• Economists: Look at job trends over time, don't focus on a single month
• Monthly revisions to past months' reports can make a big difference

A dumbfounding report Friday on job growth in September may be conveying a simple message: The labor market is in better shape than it has appeared lately.

At first blush, the report painted two starkly different pictures of the halting U.S. employment recovery. Employers added 114,000 jobs last month -- typically, not nearly enough to push down unemployment -- yet, the jobless rate fell from 8.1% to 7.8%, its lowest since Jan. 2009.

The two figures are based on separate surveys. The 114,000 gain in payrolls comes from a survey of 140,000 private and public establishments. The 7.8% jobless rate is based on a survey of 60,000 households. In that survey, the number of Americans who said they were working last month jumped by 873,000, the most since January 2003.

Which survey is telling the truth?

Economists say both actually reflect the same trend of moderate, gradually strengthening job growth when they're examined over several months. In other words, monthly variations are volatile, but three-month patterns are telling.

"I think (the surveys) are actually telling very similar stories," says Dean Maki, chief U.S. economist of Barclays Capital.

For example, the 873,000 employment increase in the household survey was preceded by a total 314,000 decline in July and August. The average monthly gain in the July-September period was 186,000.

Similarly, the tepid 114,000 increase in the establishment survey was preceded by gains of 181,000 in July and 142,000 in August, following upward revisions announced Friday by the Labor Department. The average monthly increase in the period: 146,000.

That means the establishment survey is still showing somewhat weaker job growth. But economists say the household survey more accurately reflects real-time trends in the job market, largely because it better captures new businesses and the self-employed. The establishment survey, by contrast, uses a computer model that undercounts business start-ups in a recovery and closures in a recession.

Last week, for instance, Labor tentatively said job gains in the establishment survey from April 2012 to March 2012 were 386,000 higher than it previously estimated. As a result, average monthly job gains in both surveys during that period are almost identical at about 190,000.

After warm winter weather inflated payroll gains early this year and deflated them in the spring, job growth has resumed a moderate,if unspectacular pace of about 160,000 a month, Maki says. That, he says, is more than enough to lower the unemployment rate with Baby Boomer retirements restricting growth in the number of people looking for work.

Businesses remain hesitant to hire because of concerns about Congress' ability to head off massive tax increases and spending cuts at year's end. Diane Swonk, chief economist of Mesirow Financial, notes that vehicle sales are strong, the housing market is picking up and rising home prices are making Americans feel wealthier.

"We've still got a dicey roller-coaster period," she says. But, "it looks like (the job market) is starting to improve a little bit."