In the two methods, the average income is lower, and so is the standard deviation. Which is understandable since a risk reduction involves a lower income.
Note that the diversification of cropping patterns is more significant with the expected utility maximization method. But this higher diversification level does not necessarily lead to the minimization of the standard deviation. Indeed, the standard deviation of the Target MOTAD method is the lowest. This result can be linked to the fact that even if the cropping pattern is more diversified in terms of different crops grown, two crops are hardly ever grown and the seed cabbage, for instance, which shows significant variation, is still very present. However, the four crops of the Target MOTAD model are more heterogenous.
If you did not manage to write the model or in order to check that your equations are accurate, compare with the following model : modelEco_lauragaisCARA.gms.
Case with no risk aversion | Risk with CARA utility function | Risk Target-MOTAD | |
Cropping pattern | 16 ha of durum wheat 108 ha of sunflower 100 ha of seed cabbage |
100 ha of durum wheat 49 ha of seed maize 9 ha of seed sugar beet 2 ha of seed carrot 41 ha of seed cabbage |
100 ha of durum wheat 48 ha of seed cucumber 17 ha of seed cucumber 35 ha of seed cabbage |
Land rental | 24 ha | 0 | 0 |
Average income | 469 320 | 359 029 | 360 495 |
Minimum incomes (state of nature) | -732 029 (E7) -332 199 (E10) |
57 964 (E7) 59 295 (E1) |
60 000 (E1) 110 000 (E7) |
Standard deviation | 650 654 | 359 029 | 221 934 |
Hired labour | 408 hours | 459 hours | 416 hours |