Technology and Unemployment: Empirical Evidence from System GMM Estimates

The study concentrates on whether technological progress causes job losses and unemployment. While technological progress, embodied in product and process innovation, creates new jobs, it also changes the qualifications required by existing jobs and causes machines to substitute human labour. The emergence of new jobs and productivity gains are opportunities, but whether these opportunities are sufficient to compensate for jobs lost and the resulting labour surplus is disputable. The uncertainty of the ultimate impact, methodological developments, and diversification of data sets are essential motivations for addressing the issue with different dimensions. In this study, analyses are conducted within the framework of a dynamic panel model considering the effect of the past value of the unemployment rate on its future value for the period 1995-2021 and 91 countries. The models are estimated using the two-step system Generalised Method of Moments (2SGMM). The findings of the study confirm the persistence of unemployment. Moreover, the embodied process included in gross fixed capital formation points to the labour-saving nature of innovation. Under alternative models, the economic complexity index, a measure of product innovation, is found to reduce unemployment. The number of patents included in the analysis to represent product innovation within the scope of another model affects unemployment in lagged terms. Accordingly, the increased number of patents results in a lagged decrease in unemployment rates.