You read that the latest U.S. Unemployment Rate calculation has fallen from 9.2 to 9.1 percent, but how is it calculated, who calculates it, and what does this bare statistic actually tell us?
The official Unemployment Rate calculation (UR) is performed monthly by the Bureau of Labor Statistics (BLS), a subdivision of the U.S. Department of Labor. Although relatively small as federal agencies go, the BLS enjoyed a 2011 budget of $610.2 million to fund the activities of its roughly 2500 full-time employees. The BLS’s main tool in conducting the UR is the monthly Current Population Survey (CPS), which is actually performed by 2200 employees of the U.S. Census Bureau (CB), which by comparison boasts 5,500-plus permanent employees and an annual budget averaging $3.87 billion from 2009 through 2011.
The CPS, which samples approximately 60,000 households, is compiled largely through direct interviews, and as with all surveys conducted by statistical sampling, it does not pretend to 100-percent accuracy. Instead, the CB and BLS claim a 90-percent chance that the estimate of unemployment derived from the survey is within 290,000 of the actual number of total unemployed workers.
The CPS underwent a significant redesign in 1994 which increased the overall data obtained and modified key definitions, such as that by which “discouraged workers” are described, elevating that classification to those who have (1) looked for a job within the last year and who are (2) currently available to work, as opposed to those who are “unemployed,” defined as those who (1) have looked for work in the prior four weeks, and who (2) are also currently available for work.
Given these definitions, the CPS reported total employment for July 2011 of roughly 140,000,000 workers and a UR of 9.1 percent, or approximately 13.9 million unemployed. Compare that to the BLS’s Current Employment Statistics (CES) analysis, a survey of businesses and government agencies designed to capture employment data from the perspective of employers, which reported a total employment of approximately 131,000,000 during the same month. Why the substantial difference in the two BLS counts of the employed? It is explained in part by who is surveyed and how classifications such as “employed” are defined. For example, the CPS definition of “employed” includes those who work part time in family businesses, whether or not they are paid. So, the 16 year old student who helps part time on the family farm, but who is not paid and who has no other employment, is nevertheless counted as “employed.”
The CPS and CES both offer unique strengths and weaknesses, but the CES method of identifying the “employed” appears more consistent with common notions of what it means to be gainfully employed. Comparing the CPS July 2011 total labor force tally of 154,000,000 to the CES’s total-employed count of 131,000,000 yields a percentage of unemployed that equals 14.4 percent, the statistic that many economists view as more truly descriptive of the current rate of U.S. unemployment.
In summation, the 9.1 percent UR derived from the CPS achieves its more modest statistic by counting as employed all who are arguably so, while an unofficial unemployment rate calculation of 14.4 percent is obtained if the CES’s more conservative data is substituted.