Start Date
11-12-2016 12:00 AM
Description
With the rapid advance of digital technologies, task automation has recently come to the forefront of the debate on skill-biased technological change. Building on a network theory, this study develops a new systematic methodology to identify comprehensive task types in the overall economy, and to quantitatively measure the degree of automation for each task type. Using comprehensive dataset on occupational skill requirements in 2015, we construct a two-mode network, and identify 13 task types using a non-parametric clustering algorithm. Our findings suggest that routine cognitive task and information processing are most automated tasks, and that flexible thinking and dynamic physical task are least susceptible to automation in 2015. The major contribution of our approach lies in the estimation of degree of automation for different task types. The methodology presents a promising avenue for evaluating the impact of automation on labor market outcomes, such as wage inequality and job polarization.
Recommended Citation
Park, Jiyong; Kim, Jongho; and Lee, Byungtae, "Which Tasks Will Technology Take? A New Systematic Methodology to Measure Task Automation" (2016). ICIS 2016 Proceedings. 8.
https://aisel.aisnet.org/icis2016/Sustainability/Presentations/8
Which Tasks Will Technology Take? A New Systematic Methodology to Measure Task Automation
With the rapid advance of digital technologies, task automation has recently come to the forefront of the debate on skill-biased technological change. Building on a network theory, this study develops a new systematic methodology to identify comprehensive task types in the overall economy, and to quantitatively measure the degree of automation for each task type. Using comprehensive dataset on occupational skill requirements in 2015, we construct a two-mode network, and identify 13 task types using a non-parametric clustering algorithm. Our findings suggest that routine cognitive task and information processing are most automated tasks, and that flexible thinking and dynamic physical task are least susceptible to automation in 2015. The major contribution of our approach lies in the estimation of degree of automation for different task types. The methodology presents a promising avenue for evaluating the impact of automation on labor market outcomes, such as wage inequality and job polarization.