On Production Function Estimation with Selectivity and Risk Considerations

In the estimation of production functions, ignoring risk considerations can cause inefficient estimates, while biased parameter estimates arise in the presence of sample selection. In the presence of uncertainty and selection bias, the latter introduced by the endogeneity of qualitative characteristics of inputs in crop choice, we show that correcting for risk considerations (a la Just and Pope, 1978, 1979) but not selection bias, can produce incorrect inferences in terms of risk behavior. The arguments raised in this study have estimation and policy implications for stochastic production analysis applied to all goods whose qualitative characteristics can affect sample selection.


Koundouri, Phoebe; Nauges, Celine, On Production Function Estimation with Selectivity and Risk Considerations, Journal of Agricultural and Resource Economics, Volume 30, Issue 3, December 2005, Pages 597–608

Share on twitter
Share on linkedin
Share on facebook