Munich Personal RePEc Archive

A Poisson Stochastic Frontier Model with Finite Mixture Structure

Drivas, Kyriakos and Economidou, Claire and Tsionas, Efthymios G. (2014): A Poisson Stochastic Frontier Model with Finite Mixture Structure.


Download (318kB) | Preview


Standard stochastic frontier models estimate log-linear specifications of production technology, represented mostly by production, cost, profit, revenue, and distance frontiers. We develop a methodology for stochastic frontier models of count data allowing for technological and inefficiency induced heterogeneity in the data and endogenous regressors. We derive the corresponding log-likelihood function and conditional mean of inefficiency to estimate technology regime-specific inefficiency. We further provide empirical evidence that demonstrates the applicability of the proposed model.

MPRA is a RePEc service hosted by
the Munich University Library in Germany.