On the Employment Effect of Technology: Evidence from U.S. Manufacturing for 1958-1996

Recently, Galí and others have found that technological progress may be contractionary: a favorable technology shock reduces hours worked in the short run. We ask whether this observation is robust in disaggregate data. According to our VAR analysis of 458 four-digit U.S. manufacturing industries for 1958-1996, some industries do exhibit temporary reduction in hours in response to a permanent increase in TFP. However, there are far more industries in which technological progress significantly increases hours. Using micro data on average price duration, we ask whether the difference across industries is related to the stickiness of industry-output prices. Among 87 manufacturing goods, we do not find such a relation.