Empirical Essays on the Elasticity of Substitution, Technical Change, and Economic Growth

Abstract

We estimate the elasticity of substitution using two different production functions. The usual Constant Elasticity of Substitution (CES) production function and a Box-Cox production function for Japan (1890-1991), UK (1870-1991), and US (1890-1992; 1929-2000). The main results are that we find the ES to be non-unitary and changing over time. Our findings have implications for economic growth (theoretical and empirical), as production is an increasing function of the ES. The use of a Cobb-Douglas production function, as in most cases in the literature, hides the role of the ES not only as a source of increase in output but also as a source of technical change. Also, we found in a monte carlo simulation that usual CES production usually does not give reliable estimates for the substitution parameter. Finally, using a CES to calculate TFP across countries, we found that variance of TFP is lower than using a Cobb-Douglas. It implies that the importance of TFP to explain income differences across countries is diminished.