Distributions Of Public And Private Manufacturing Firms And Determinants Of Productivity In Ethiopia

The aim of this paper is to assess the structure and distribution of large and medium (L&M) scale manufacturing industries in Ethiopia and to investigate the level of their productivities by disaggregating them into public and private firms using firm level panel data (from 2003 to 2005) for manufacturing industries. The paper employs Cobb-Douglas type production function estimated using three alternative techniques: Ordinary Least Squares (OLS), Fixed Effect (FE) and Generalized Method of Moments (GMM). The GMM results are found to be unbiased estimates for our dataset. Considerable regional variations are observed in the distribution of firms in Ethiopia, more so after 1991 than in the period before it. Significant regional variations are also observed in terms of production capacity and capital intensity. Compared to their counterparts, public firms employ substantially higher level of factors inputs and hence produce equivalently larger amount of output. Nevertheless, the GMM estimations reveal that despite public firms’ substantial access to factor inputs, no statistically significant productivity differential is observed between the two sectors. Finally, while only indirect and material inputs are found to significantly influence publicly owned firms’ productivities, all of the model variables (physical and human capital, indirect and material inputs, and workers’ experience) are found to determine the productivities of private firm to a great extent.