Estimation of Technical Change and TFP Growth Based on Observable Technology Shifters

Abstract

This paper models and estimates total factor productivity (TFP) growth parametrically. The model is a generalization of the traditional production model where technology is represented by a time trend. TFP growth is decomposed into unobservable technical change, scale economies and observable technology shifter index components. The empirical results are based on an unbalanced panel data at the global level for 190 countries observed over the period 1996-2013. A number of exogenous growth factors are used in modeling four technology shifter indices to explore development infrastructure, finances, technology and human development determinants of TFP growth. Our results show that unobservable technical changes remain the most important component of TFP growth. The observable technology indices-based component is lower than the simple unobserved time trend model based one. By comparing the performance of the time trend and technology index models in terms of TFP growth rates, we arrive at the conclusion that the technology index model predicts a more realistic picture of the TFP growth pattern as compared to the traditional time trend model. Our results also indicate that technical change and TFP growth are negative across country groups and years in the technology index model influenced by the global economic crisis.