Average Work Hours as a Leading Economic Variable in U.S. Manufacturing

**Below is a description of the paper and not the actual abstract.** This paper examines the role of the average workweek in U.S. manufacturing industries as a leading indicator. Our analysis investigates the relationship between average weekly hours of production workers in manufacturing, a leading indicator, and both employment of production workers in manufacturing and manufacturing output, which are coincident indicators. A separate VAR system hours, employment, output and real earnings is estimated for the aggregate, durable and non- durable manufacturing sectors as well as for sixteen of the twenty 2-digit SIC manufacturing industries. Tests for structural change show a structural break occurred after 1978. Granger causality tests and impulse response analysis both show that the impact of average and overtime hours on employment and industrial output became considerably weaker after the structural break point. In the period 1979 to 1993, considerably fewer detailed industries exhibit significant impulse responses of employment and output to a unit shock in hours as compared to the period 1954 to 1978. Our results imply that a given change in hours is now leading to smaller changes in subsequent employment levels. In general, the average workweek length in U.S. manufacturing has become less associated with the entire business cycle