Comment on “Computerizing industries and routinizing jobs: Explaining trends in aggregate productivity” by Sangmin Aum, Sang Yoon (Tim) Lee and Yongseok Shin

Aum et al. (2018) quantify the impact of production complementarities and differential productivity growth across occupations and sectors on the slowdown of aggregate productivity growth. This note expands their work to study substitutability between new computer equipment and labor in individual occupations as opposed to all occupations combined. Preliminary empirical evidence suggests (1) significantly different elasticities of substitution between computers and labor across occupations and (2) a strong correlation between productivity growth of computers and labor in occupations where these two inputs are complementary. When they are substitutes, however, their productivity growth rates appear uncorrelated. These findings have the potential to amplify or weaken the magnitude of the aggregate productivity slowdown explained by Aum et al. (2018) making their approach a promising avenue for future research.

Duke Scholars

Author Matthias Kehrig Economics