Modelling labour productivity and the role of research intensity in 129 years: evidence from a new dynamic instrumental variable estimation approach

One of the areas of empirical research that has remained largely underexplored is the role of research intensity on labour productivity. To bridge this gap, this study investigates the impact of research intensity on labour productivity in 15 OECD countries using a new dynamic instrumental variable estimation approach while providing for financial development and education level as additional control variables for the 1890–2018 period. The results reveal that in most cases, research intensity, financial development, and level of education have positive impacts on labour productivity. In the model containing all the control variables and dummies, it is observed that for every 1 percentage point increase in research intensity, the labour productivity grows by 0.614 points. The results of the dummies suggest mixed evidence for the impact of World War I on labour productivity. The results further suggest that both World War II and the oil price crisis of the 1970s led to positive changes in labour productivity. For the purpose of robustness, we have also used a new Poisson pseudo maximum likelihood estimation approach to estimate the impact of research intensity on labour productivity, but the results are not materially different. The policy implications of the empirical results are detailed in the paper.