Automation and New Tasks - How Technology Displaces and Reinstates Labor Figures

Labor share, 1947−1987 80 70 Manufacturing Construction 60 50 Services Transportation 40 Mining 30 Agriculture 20 1947 1952 1957 1962 1967 1972 1977 1982 1987 Share GDP, 1947−1987 Services 60 50 40 30 Manufacturing 20 10 Agriculture 0 1947 1952 1957 1962 1967 1972 1977 1982 Transportation Construction Mining 1987 Figure 2: The labor share and sectoral evolutions, 1947-1987. Note: The top panel shows the labor share in value added in services, manufacturing, construction, transportation, mining and agriculture between 1947 and 1987, while the bottom panel shows the share of value added in these sectors relative to GDP. Data from the BEA industry accounts. Wage bill, 1947−1987 100 Observed wage bill Productivity effect 80 60 40 20 Substitution effect Composition effect Change in task content 0 1947 1952 1957 1962 1967 1972 1977 1982 1987 Change in task content of production, 1947−1987 20 Reinstatement 10 Change in task content 0 −10 Displacement −20 1947 1952 1957 1962 1967 1972 1977 1982 1987 Manufacturing task content of production, 1947−1987 20 Reinstatement 10 Change in task content 0 −10 −20 1947 Displacement 1952 1957 1962 1967 1972 1977 1982 1987 Figure 3: Sources of changes in labor demand, 1947-1987. Note: The top panel presents the decomposition of wage bill divided by population between 1947 and 1987. The middle and bottom panels present our estimates of the displacement and reinstatement effects for the entire economy and the manufacturing sector, respectively. See text for the details of the estimation of the changes in task content and displacement and reinstatement effects. Labor share, 1987−2017 70 Construction 60 Services Transportation Manufacturing 50 40 Agriculture Mining 30 20 1987 1992 1997 2002 2007 2012 2017 Share GDP, 1987−2017 Services 70 60 50 40 30 20 Manufacturing 10 0 1987 Agriculture 1992 1997 2002 2007 2012 Transportation Construction Mining 2017 Figure 4: The labor share and sectoral evolutions, 1987-2017. Note: The top panel shows the labor share in value added in services, manufacturing, construction, transportation, mining and agriculture between 1987 and 2017, while the bottom panel shows the share of value added in these sectors relative to GDP. Data from the BEA industry accounts. Wage bill, 1987−2017 Productivity effect 40 Observed wage bill 30 20 10 Composition effect Substitution effect 0 Change in task content −10 1987 1992 1997 2002 2007 2012 2017 Change in task content of production, 1987−2017 Reinstatement 10 5 0 −5 −10 Change in task content −15 −20 1987 Displacement 1992 1997 2002 2007 2012 2017 Manufacturing task content of production, 1987−2017 Reinstatement 10 0 −10 Change in task content −20 −30 Displacement 1987 1992 1997 2002 2007 2012 2017 Figure 5: Sources of changes in labor demand, 1987-2017. Note: The top panel presents the decomposition of wage bill divided by population between 1987 and 2017. The middle and bottom panels present our estimates of the displacement and reinstatement effects for the entire economy and the manufacturing sector, respectively. See text for the details of the estimation of the changes in task content and displacement and reinstatement effects. Table 1: Relationship between change in task content of production and proxies for automation and new tasks. (1) Controlling for manufacturing (2) Controlling for Chinese import and offshoring (3) -1.404 (0.377) 61 0.18 -0.985 (0.369) 61 0.21 -1.129 (0.362) 61 0.27 Observations R-squared -0.394 (0.122) 61 0.14 -0.241 (0.159) 61 0.19 -0.321 (0.164) 61 0.27 Detailed manufacturing industries (SMT): Share of firms using automation technologies, 1988-1993 Observations R-squared -0.390 (0.165) 148 0.08 Proxies for new tasks: Share of new job titles, based on 1991 DOT and 1990 employment by occupation Observations R-squared 1.609 (0.523) 61 0.12 1.336 (0.530) 61 0.23 1.602 (0.541) 61 0.32 Number of emerging tasks, based on 1990 employment by occupation Observations R-squared 8.423 (2.261) 61 0.14 7.108 (2.366) 61 0.25 7.728 (2.418) 61 0.33 Share of growth between 1990-2016 in occupations not in industry in 1990 Observations R-squared 2.121 (0.723) 61 0.08 1.638 (0.669) 61 0.20 1.646 (0.679) 61 0.26 Percent increase in number of occupations represented in industry Observations R-squared 0.585 (0.156) 61 0.14 0.368 (0.207) 61 0.19 0.351 (0.215) 61 0.25 Raw data Proxies for automation technologies: Adjusted penetration of robots, 1993-2014 Observations R-squared Share of routine jobs in industry, 1990 -0.397 (0.166) 148 0.09 Note: The table reports estimates of the relationship between the change in task content of production between 1987-2017 and proxies for automation technologies and new tasks. Column 1 reports estimates of the bivariate relationship between change in task content of production and the indicated proxy at the industry level. Column 2 includes a dummy for manufacturing industries as a control. In addition, Column 3 controls for the increase in Chinese imports (defined as the increase in imports relative to US consumption between 1991 and 2011, as in Acemoglu et al. 2016) and the increase in offshoring (defined as the increase in the share of imported intermediates between 1993 and 2007, as in Feenstra and Hanson, 1999). Except for the panels using the Survey of Manufacturing Technologies (SMT), all regressions are for the 61 industries used in or analysis of the 1987-2017 period. When using the SMT, the regressions are for 148 detailed manufacturing industries. Standard errors robust against heteroskedasticity are in parenthesis. When using the measure of robot penetration, we cluster standard errors at the 19 industries for which this measure is available.