Multicrop Production Models with Multinomial Logit Acreage Shares

The main objective of this paper is to show that the MultiNomial Logit (MNL) framework can be used for deriving simple acreage share models. More specifically we show that MNL acreage choice models can be defined as parts of consistent multicrop models. When coupled with suitably designed crop production models, e.g. a specific parameterization of the standard quadratic yield function, the parameters of these multicrop models are easily interpreted and estimated. Although MNL acreage choice models are usually derived by aggregating crop choices defined at the plot level, we show that these models can also be derived by considering a profit maximisation problem at the farm level. This provides a simple theoretical background for the Standard MNL acreage share model, as well as for simple extensions of these models: the Nested MNL acreage share models. An empirical application on French farm-level data over the period 1996–2007 investigates the empirical performances of the multicrop econometric models based on the proposed MNL framework for modelling acreage choices. Copyright Springer Science+Business Media Dordrecht 2014