Total Factor Productivity (TFP)is often used on the macro-economic level as an indicator of changes in efficiency of a country. In many transition economies TFP is seen to have been negative the last decade of the plan economy and starts increasing and become positive after a (quite a) few years of transition. Many authors conclude that this is a gain in efficiency due to the structural changes –such as privatisation and liberalisation – carried out in order to establish a market economy in those countries. In the case of Mongolia, not only non-viable enterprises closed down, but many possibly viable enterprises with potential closed down as well. This raises the question whether changes in TFP were really attributable to increases in efficiency. To investigate this, the mathematical properties of TFP are analysed in order to generate new insights into the development of TFP in Mongolia. Simulations are performed to see what happens with TFP if not the le! ast efficient, but a certain percentage of enterprises in a (closed) economy randomly close down. The robustness of Total Factor Productivity of Mongolia was tested not only for errors in all estimated values but also for measurement errors in the data. It was concluded that in many commonly occurring cases it is not necessary to estimate alpha; that a random closure of enterprises fits the data of Mongolia much more closely than closing only the least efficient enterprises; and that measurement errors in the data influence the estimated TFP significantly.
transition development TFP total factor productivity Mongolia measurement errors simulation Cobb-Douglas production function sensitivity analysis efficiency