PDF Ebook A mixture based stochastic frontier model
The analysis of the production/cost functions is a central theme in the econometric literature. In particular, many are the works that develop efficiency evaluation procedure connected with the production and cost “frontiers” estimation. The present study focuses on the problem of the technical efficiency measurement but the proposed approach can be easily adapted to the dual concept of economic efficiency developed in the seminal work of Farrell (1957) and the following papers on cost efficiency analysis (see, for instance, Koop and Diewert, 1982).
In a fully efficient framework the production/cost functions can be estimated considering the classical statistical regression theory. Full efficiency is connected with the presence of the so called “best practitioners”. In a market based economy all the productive units are thought to operate in efficient framework otherwise they are automatically out of business. In the real data context, even if the market regulates the demand/supply dynamics, a subset of the productive units would, certainly, operate inefficiently. This is the motivating idea for the papers of Debreu (1951) and Koopmans (1951) on the measurement of the productive efficiency. In particular, a productive unit is said to be technical efficient if, given a fixed amount of resources, it maximizes the output.
The application of econometric approaches in this context goes back to the early studies of Aigner and Chu (1968), Aigner et al. (1977) and Meeusen and van den Broek (1977) on the estimation of the so called “frontier production function”. In the parametric models context one can distinguish between deterministic (corrected and modified OLS procedures Aigner et al., 1977; Greene, 1980a; Stevenson, 1980; Greene, 2003) and stochastic approaches (Aigner et al., 1977; Greene, 1982, 1990a; Kumbhakar and Lovell, 2000). Inparticular, the works of Aigner et al. (1977) and Meeusen and van den Broek (1977) introduce the concept of stochastic frontier (SF) model specification. The model generalization to panel data treatment has been largely studied, in particular, considering the SF approach (see, for example, Hausman and Schmidt, 1981; Schmidt and Sickles, 1984; Battese and Coelli, 1988, 1992; Fernandez et al., 1997; Kumbhakar and Lovell, 2000).
Download
PDF Ebook A mixture based stochastic frontier model
Posted in :
