I am using a panel data set for the manufacturing industries (firm-level data) for about 20 years, which includes 4 inputs (Capital, labour, energy and materials) and 1 output (sales revenue). I am going to use a stochastic frontier analysis (Translog input distance function approach) to calculate the elasticity of substitution (between factors of production) and technical efficiency (time-varying model preferably with half normal / truncated normal distribution) using the Maximum likelihood estimation (MLE).
I wish to seek some help regarding cleaning my data and subsequently formulating my STATA do-file for the project as well. My key objectives include:
1. Translog input distance function need to formulated basis Coelli and Perelman (1995,1999)
2. Technical efficiency needs to calculated basis Battese and Coelli (1992,1995) and further determinants of the technical efficiency also need to be evaluated basis a second stage regression.
3. The elasticity of substitution need to be calculated for multi-input case (basis Blackorby and Russell 1989, i.e. Allen elasticity of substitution (AES), Morishima elasticity of substitution (MES), and the shadow elasticity of substitution (SES)
4. The direction of the causality of the relationship between technical efficiency and the elasticity of substitution need to be evaluated. The direction of this auxiliary regression will suggest as to whether technical efficiency is impacted by the elasticity of substitution or vice versa?
5. With my final objective is to seek the responsiveness of elasticity of substitution and technical efficiency (i.e. both separately) with respect to the varying firm sizes (in terms of employee headcount) and sub-sectoral classification (NAICS-4 level i.e. focussing on the sub-sectors of the food manufacturing industry) of the manufacturing industry.
A brief background/ Intuition of the project: This economics project seeks to explain the impact of the rising input cost as the factors of production (capital, labour, energy and materials) in the manufacturing industry are getting expensive. So in order to minimize the cost, the firms would like to substitute from expensive to cheaper inputs but without compromising on the level of output generated. The rise in the input cost might also hit the technical efficiency of the firms as well and the extent to which stands to reason basis our results. We are proposing to use the translog input distance function to analyze (which is dual to the cost function), and not the cost function because we don't have access to the prices of inputs.
Hope it helps with the intuition part if more information desired then please contact.
10 freelancere byder i gennemsnit $42 på dette job
I am using a panel data set for the manufacturing industries (firm-level data) for about 20 years, which includes 4 inputs (Capital, labour, energy and materials) and 1 output (sales revenue). I am go... Read More