| This study has used a corrected likelihood ratio, with AIDS and Rotterdam demand models, to test separability in three separable (A, B, C) wheat import demand structures on the French common wheat market. It appears from the study that the three separable Rotterdam structures are accepted by the test, while only two of them (B and C) are accepted in the case of AIDS at 5% level of significance. Meanwhile, model B seems to be more indicated in demand analysis with AIDS since it is the only one accepted at 15% level of significance. The results obtained demonstrate not only the necessity to test separability in demand structures, but also the necessity to know how the allocation models can be used for demand analysis purpose. The study has also shown a great... |
