Estimation of the elasticity of substitution of production factors in CEE economies

I estimate the elasticity of substitution between capital, labour and energy in Central and Eastern European countries to make CGE models more accurate. The CGE models often use elasticity of substitution of production factors estimated for another countries or average of elasticities taken from literature. I want to provide new estimates of elasticity of substitution of production factors for CGE models. I focus on CEE countries where no estimates of elasticity of substitution of production factors have been done yet. The CEE economies differ from the Western European economies therefore we have reasonable argument to suppose that also the elasticities of substitution of production factors are different in these countries. The newly estimated elasticities of production can be used as parameters for the CGE models and enable comparison of the results, when using the newly estimated parameters and when using the conventional value of parameters. The need to estimate constant elasticity of substitution (CES) production function with nested structure is discussed since Berndt and Wood (1979). In 2-level nested ((KE)(LM)) production function, Berndt and Wood (1979) distinguish the gross substitution effect and the expansion effect. They explain that the engineering studies supporting the E-K substitutability capture only the gross substitution effect and therefore these studies might not be in contradiction with econometric studies supporting the E-K complementarity, because the expansion effect could outweigh the gross substitution effect. Kemfert (1998) estimates the substitution elasticities of a nested CES production function for the entire German industry and individual industrial sectors. She estimates the elasticities for tree nested structures [((KE)L), ((KL)E) and ((EL)K)] and concludes that a nested CES function with a nest of capital and energy ((KE)L) is the most useful for the entire German industry, but for several industrial sectors a nest of capital and labour ((KL)E) might be closer to the reality. Van der Werf (2008) estimates a 2-level nested production function, using the industrial level data from 12 OECD countries in for 1978-1996. He finds that the nesting structure having capital and labour in the same node fits reality most closely. Similarly, Okagawa and Ban (2008) estimate a nested CES function using another OECD dataset. Their data set is more refined compared to that used in Van der Werf (2008), where the data are disaggregated into 7 sectors; the Okagawa and Ban (2008) data set disaggregated into 19 sectors. Following these findings I estimate a production function with a nested structure and examine, which nested structure fits data best. I use the World Input-Output Database (WIOD) which includes data from 27 EU countries plus 13 other major countries for the period from 1995 to 2009. The database distinguishes 35 industries, so I estimate sector specific elasticities for each CEE country from the EU. New estimates of elasticity of substitution between capital, labour and energy in Central and Eastern European countries that can be used as parameters in CGE models.