We had a number of inquiries about our unemployment post yesterday. A lot of readers asked us why we seemed to be suggesting that US Job losses were decelerating while at the same time the unemployment rate was accelerating , and how the math behind that observation worked.
The math is actually very straightforward: This divergence between measures of employment stems from differences in the methodology used by the government to calculate different figures, combined with cultural and psychological inflections.
The Current Population Survey (the monthly survey of 60,000 US households used to calculate the unemployment rate) classifies people in one of three ways: "Employed", "Unemployed" or "Not in the Labor Force". Unemployed people are defined as: jobless, looking for jobs, and available for work. This means that those not looking for work and not available for work are counted as "Not in the Labor Force" even if they have no current source of income. Meanwhile, payroll data is collected through a business survey of wages paid and Initial Claims are calculated through a count of new state benefits applications.
During periods of increasing layoffs, a phenomenon sometimes occurs in which the first wave of individuals who lose their jobs do not feel compelled to begin looking for new employment until after successive waves of additional individuals entering the market has created greater competition for available positions. I'll use myself as an example: In early 2001 when I was laid off by an investment bank I decided to kick back with my (by the standards of 29 year old) considerable savings and take the summer off while I recharged the batteries and considered my next move.
As such, an initial wave of unemployed workers like my 29 year old self are sometimes less likely to identify themselves as actively seeking employment immediately than people who lose their jobs at a later time will, and those initial individuals respond to increasing competition later by re-entering the market and identifying themselves as seeking a job. I'll use myself as an example again, as the NYC finance job situation deteriorated post 9/11; I was suddenly energized in my job search and began looking aggressively.
In the charts below you can see how payroll data began to show improvement while the unemployment rate was still climbing during periods of increased job loss during 1990-92 and 2001-03 and how the current cycle seems to be indicating a similar pattern (at an accelerated rate and trajectory).