Substitution Elasticities in the Large-Scale Manufacturing Industries of Pakistan--a Rejoinder (Communication) (Report‪)‬

Estimates of the elasticity of substitution between capital and labour in 16 manufacturing industries and the manufacturing sector of Pakistan as a whole were reported in [18]. In most of the industries and the manufacturing sector as a whole substitution elasticities were found to be rather low. In the Spring 1982 issue of this Review, Ahmed [1] has suggested that the estimates of substitution elasticities may have been biased downwards due to the irrelevance of production functions in the developing countries and the nature of data employed in the study. He also believes that the evidence we presented regarding low substitution elasticities in the other developing countries was selective because according to him Morawetz [23] provides evidence to the contrary. Furthermore, he argues that low substitution elasticities are inconsistent with the declining share of labour in the output. In the following, without being drawn into polemics, we shall show that Ahmed's comments are a result of his misunderstanding and misinterpretation of our study. Broadly speaking, Ahmed's comments may be divided into four parts: (i) conceptual and estimation problems relating to production functions; (ii) nature of data used in our study and their effect on the estimates of substitution elasticities; (iii) comparison of the substitution elasticities in Pakistan with those in other developing countries; and (iv) problems relating to interpretation of results and drawing of policy implications from the analysis. These comments are taken up in this same order.