Are Robots, Software, ICT and physical capital related to productivity? A panel quantile approach

• Published in: Economics of innovation and new technology Vol. 33; no. 4; pp. 586 - 603
• Main Authors: Almeida, Derick, Neves Sequeira, Tiago
• Format: Journal Article
• Language: English
• Published: Abingdon Routledge 18-05-2024; Taylor & Francis Ltd
• Subjects: Capital; Communications technology; Elasticity; Elasticity of demand; Information technology; Labor; Labor productivity; panel data; Productivity; quantile regression; Quantiles; Regression analysis; Robotics; Robots; Software; wage inequality
• ISSN: 1043-8599; 1476-8364
• DOI: 10.1080/10438599.2023.2227954

We study comparable elasticities of labor productivity on Robots, ICT, Non-ICT and Software for the whole conditional distribution of labor productivity through quantile regression analysis. When considering both country-industry fixed-effects and IV approaches, we obtain an increasing elasticity of labor productivity with respect to robot density across the main deciles of the labor productivity distribution. On average, a 1% increase in robot density is associated with nearly 0.15% increase at the 1st decile with nearly 0.17% at the 9
th
decile of productivity, an effect that is nearly nonsignificant within manufacturing and is crucially dependent on the lower robotized sectors. The elasticity of labor productivity to non-ICT capital (about 0.4%) and software (reaching 0.23%) are shown to be higher than that towards robots. Also the elasticities towards ICT and software vary more across the productivity distribution - presenting a downward slope - than those towards robots and physical capital. This may indicate an association of ICT and software technologies with decreasing inequality and of Robots with increasing inequality.