Are Industrial Robots a new GPT? A Panel Study of Nine European Countries with Capital and Quality-adjusted Industrial Robots as Drivers of Labour Productivity Growth

In recent years, the interest in the field of economic research in studying the effect of robots on economic outcomes, i.e., labour productivity, labour demand and wages, has increased from an individual country perspective as well as for country groups. By using a fixed effects panel modeling approach, this study of nine robot intensive European countries shows that the core characteristics of a general purpose technology (GPT) are already satisfied by industrial robots. In 2019, seven countries in the panel, i.e. Germany, Italy, France, Spain and the UK (top 5), Sweden (7th) and Austria (10th) - in terms of operational stocks - were among the top 10 of robot using European countries (excl. Turkey). Following the understanding of a GPT of Bresnahan/Trajtenberg (1995), six panel regression models were estimated and linked to the four main characteristics of a GPT. Accordingly, two new measures are proposed in this paper; the first one is named the Division of Labour (or DoL) and is constructed by building the ratio of labour productivity inside the manufacturing industry to labour productivity across all industries. The second one is the Robot Task Intensity Index (RTII), which accounts for the number of tasks that a robot was used for in different production processes across the nine European countries. A high level of fulfilled tasks implies a higher quality of robot as the number of potential tasks, which the robot can perform, is an important criterion for the quality of that robot. In accordance with the GPT literature, both measures showed the expected (in) significances. At the bottom line, all six models underlined the economic relevance of industrial robots for the nine European countries included in the analysis and give a strong indication that robots can indeed be seen as a new general purpose technology.