Underemployment Improves Slightly in October - Proposed New Measure U7 Falls from 13.0% to 12.8%

The U.S. unemployment rate (U3) declined slightly to 9.0 percent, not much comfort. But if we add the percentage of the labor force that is underemployed –the involuntarily employed part-time, i.e., those working “part-time for economic reasons” – the combined measure of unemployment fell by two-tenths of a percentage point, from 13.0 to 12.8 percent. This is an unofficial but useful number that we might call U7. See the last line of the following table.

Un- and Under-Employment Indicators	Aug	Sep	Oct
Unemployed/labor force, Rate % (U3)	9.1	9.1	9.0
Labor Force (mil.)	153.6	154.0	154.2
Unemployed (mil.)	14.0	14.0	14.0
Teenagers unemployed %	25.4	24.6	24.1
Blacks unemployed %	16.7	16.0	15.1
Long-term unemployed/unemployed %	42.9	44.6	42.4
Unemployed+discouraged/LF+disc* % (U4)	9.7	9.7	9.6
Unemployed + marg. att./labor force+marg. att.* % (U5)	10.6	10.5	10.5
Unemployed + involuntarily PT/LF % (proposed U7)	12.9	13.0	12.8

*Not seasonally adjusted and therefore should not be combined with (non-starred) seasonally adjusted data. Marginally attached includes discouraged workers. Denominator for all indicators except the long-term unemployed is the labor force. Seasonal adjustments are by the BLS. Source: Department of Labor, Bureau of Labor Statistics, www.bls.gov/news.release/pdf/empsit.pdf, November 4, 2011. The official broad unemployment and underemployment rates (U4, U5, U6) are discussed at www.bls.gov/fls/flscomparelf/unemployment.htm#table1_2; they combine seasonally adjusted and non-seasonally adjusted numbers in the numerator and the denominator. U7 is a proposed indicator that combines two seasonally adjusted numbers and divides them by another seasonally adjusted number, avoidjing the seasonal adjustment issue and permitting month-to-month comparisons.

It is a mystery to me why the BLS uses as its only broad measures of unemployment three numbers (U4, U5, U6) that combine seasonally adjusted numbers with unadjusted numbers. This means that monthly changes in these numbers may be misleading because of seasonal factors. The proposed U7 measure does not suffer from this problem and is easily interpreted.

A consequence of combining seasonally adjusted with unadjusted numbers is that state data are provided only on a four-quarter moving average, which is pretty useless for picking up state underemployment trends on a timely basis. 

The U7 indicator (unemployment + underemployment) can be compared on a monthly basis because both components (and the labor force denominator) are seasonally adjusted.