Chapter 10: Problem 12
Many empirical studies of costs report an alternative definition of the elasticity of substitution between inputs. This alternative definition was first proposed by \(\mathrm{R}\). G. \(\mathrm{D}\). Allen in the \(1930 \mathrm{s}\) and further clarified by H. Uzawa in the 1960 s. This definition builds directly on the production function-based elasticity of substitution defined in footnote 6 of Chapter 9 : \(A_{i j}=C_{i j} C / C_{i} C_{j},\) where the subscripts indicate partial differentiation with respect to various input prices. Clearly, the Allen definition is symmetric. a. Show that \(A_{i j}=e_{x_{i}, w_{j}} / s_{j},\) where \(s_{j}\) is the share of input \(j\) in total cost. b. Show that the elasticity of \(s_{i}\) with respect to the price of input \(j\) is related to the Allen elasticity by \(e_{s_{r}, p_{j}}=s_{j}\left(A_{i j}-1\right)\) c. Show that, with only two inputs, \(A_{k l}=1\) for the CobbDouglas case and \(A_{k l}=\sigma\) for the CES case. d. Read Blackorby and Russell (1989: "Will the Real Elas- ticity of Substitution Please Stand Up?") to see why the Morishima definition is preferred for most purposes.