Even as the Bureau of Labor Statistics reported Friday the economy shed 159,000 jobs -- the largest monthly payroll decline since March 2003 -- the unemployment rate held steady at 6.1%.

The seeming inconsistency -- a loss of jobs but no increase in the unemployment rate -- flows from the fact the BLS on the first Friday of each month reports the results of two separate surveys, one measuring jobs and the other measuring the number of people working.

The jobs tally comes from the Current Employment Statistics or “establishment survey” which is a monthly sample of 400,000 businesses. The unemployment rate comes from the Current Population Survey or “household survey” which samples 60,000 households.

For both surveys, the data for a given month, according to BLS, relate to a specific week or pay period. “In the household survey,” the BLS explains in its technical note, “the reference week is generally the calendar week that contains the 12th day of the month. In the establishment survey, the reference period is the pay period including the 12^th which may or may not correspond to the calendar week.”

As odd as it may sound, jobs don’t always correspond to the number of people working. For one thing, the surveys use different definitions.

• The household survey includes agricultural workers, self-employed, unpaid family workers and private household workers, all of whom are excluded from the establishment survey.
• The household survey includes individuals on unpaid leave. The establishment survey does not.
• The household survey is limited to individuals 16 and older while there is no minimum age for the establishment survey.
• The household survey counts individuals once even if they have more than one job. Someone working at two jobs could show up twice in the establishment survey.

That is not to say the two surveys don’t influence each other.

The unemployment rate is calculated by dividing the number of people unemployed by the civilian labor force which is the sum of those employed and unemployed. But “unemployed” is a defined term. To be considered unemployed, an individual has to be out of work, available for work and looking for work. Someone who loses a job but doesn’t bother to look for work is not counted as unemployed.

In some instances, an increase in the unemployment rate can be a good sign. Consider, for example, a working age population of 11 people, nine of whom are working, one of whom is unemployed (meeting the definition) and one of whom is neither. The unemployment rate would be 10%: one unemployed divided by the sum of those employed and unemployed. Now assume the economy starts to improve. The individual on the sidelines is encouraged at the possibility of getting a job and starts to look, meeting the definition of “unemployed.” Now, the labor force totals eleven: nine employed and two unemployed producing an unemployment rate of 18%, a higher unemployment rate even as the economy improves.

That’s not the case just now as jobs are not being created. In fact, the unemployment rate jumped in August because unemployment increased. That may sound obvious, but consider what happened in August. The number of people employed fell by 342,000, but the number of people unemployed jumped by 592,000. If only employment had dropped in August -- with no change in the number of people unemployed -- the unemployment rate would have remained at 5.7%, where it had been in July.

Economists speculated in August the number of individuals unemployed jumped in August because household budgets were squeezed and indeed there is some evidence to support that contention. Of the 592,000 new “unemployed,” about 483,000 were women. The labor force participation rate -- that is, the percentage employed and unemployed as a percentage of the entire population 16 and older -- for women is generally lower than for men. That the women contributed so disproportionately to the increase in unemployment suggested they entered the labor force, as unemployed, because of the need to supplement household income.

Unemployment rose in September by about 100,000. The number of unemployed women fell by 273,000 while the number of unemployed men rose 373,000. At the same time, employment in September fell 222,000, which means not all of those who stopped working became unemployed.

Mark Lieberman is the senior economist for the Fox Business Network. Prior to joining FOX, he served as first vice president and manager of economic analysis and research at Washington Mutual in New York. Before that, he served as senior vice president at Dime Savings Bank of New York (which was later acquired by Washington Mutual), where he specialized in credit and risk management. He is a member of the Executive Committee of the New York Association for Business Economics. He has a degree in Economics from the Wharton School of the University of Pennsylvania.