Emergency aid may be inflating unemployment

WASHINGTON (Reuters) - A surprising jump in the U.S. unemployment rate last month may partly reflect temporary emergency aid that encouraged more people to count themselves as unemployed, but this ought to fade when the program ends.

The jobless rate surged to 6.1 percent in August from 5.7 percent the month before, the Labor Department said on Friday, underlining the weak state of the economy and suggesting a recession may be unavoidable.

But a chunk of this deterioration stemmed from a 250,000 increase in the civilian labor force that some economists said looked suspicious. Employment fell by 342,000 people and together, this pushed unemployment up by 592,000 last month.

"People reclassified themselves as a result of moving onto the job rolls for the supplemental program," said Lou Crandall, chief economist at Wrightson IPAC in Jersey City, New Jersey.

He estimated that without the program, which extended the duration of jobless benefits to 26 weeks from 13 weeks, unemployment would have increased to just 5.8 percent. Analysts polled by Reuters before Friday's monthly payroll report had expected it would remain unchanged.

To count as unemployed, respondents to a government survey of households who are out of work must say they have actively looked for work in the last four weeks.

People who may have been out of work for a while, but got discouraged in their job search and stopped looking, are not counted as part of the official labor force.

The number of these 'marginally attached' workers fell 200,000 last month.

"The extension of unemployment claims benefits may be causing a larger number of people than usual to state they are unemployed because this allows them to receive extended benefits," wrote Wachovia senior economist Mark Vitner.

Vitner said that the normal pattern was for the labor force to decline when the employment climate worsens. If it had not changed last month, the scale of the increase in the unemployment rate would have been halved, to 5.9 percent.

The Labor Department said 1.4 million people claimed emergency unemployment compensation in the week ended August 16, which was the week of the monthly employment survey.

"To receive benefits you must certify that you were looking for a job. So when another government agency comes around to take a survey to ask whether you've looked for a job in the last four weeks, your answer is now, 'Yes'," said Crandall.

"You only need a fraction of those to change the category they classify themselves in," he said.

A similar thing happened back in 2002, the last time there was a nationwide extension of unemployment benefits.

Crandall noted that in the first two months of that program, the jobless rate rose to 6 percent from 5.5, compared to the increase to 6.1 percent from 5.5 this time around.

"By and large what (the program) does is shift the reported unemployment rate up for the duration of the program. The same thing happened in 1994," he said.

Not everyone was convinced that the scale of the impact was sufficient to shift the basic message of a very weak economy.

Lehman Brothers economist Zach Pandl said that if the decline in the marginally attached workers was entirely due to people reclassifying themselves to get extra money, the upward bias in the unemployment rate was 0.2 percentage points.

But the numbers were volatile and there were other explanations as to why this category could decline.

"The economic story makes sense but I just don't find compelling evidence in the data," he said.

(Editing by James Dalgleish)