f(p1) =
v(CE) - v(xl)
v(x∏) - V(Xl)
(14)
The function value, that was estimated previously, doesn't need the knowledge of the
decision weights to determine its value. Substituting in the previous equation the value
function (equation 3) and given that xL = 0 in the elicitation of the certainty equivalent,
then f +(p) and f -(p) are calculated by the following expressions:
f + , f - - probability weighting function for positive and negative values;
f+(p)=
( CEI
I xι )
where:
√ω1
and f-(p)=
f ⅛2
I xι )
(15)
p - probabilities;
CE1 ,CE2 - positive and negative certainty equivalents;
x1, x1’- positive and negative results; and,
ω1, ω2 - parameters of the value function.
The probability weighting function is estimated by the confrontation of the probabilities
presented above to the decision makers with the resulting values calculated by the above
formulas. The elicitation process is independent of the value function and of the decision
weights used in this research work that was recommended by Quiggin (1993) and used by
Bouzit and Gleyses (1996) for estimating the functions of the rank-dependent Expected
Utility.
The answer to the first objective, that seeks to characterize the farmers' behavior undre
the mid-term review of the Common Agricultural Policy, will be achieved by the
adjustment of the mathematical programming model to each farmer decisions. Here, the
value of the objective function represents the farm income in each nature state. The
second objective studies the introduction of the area-yield crop insurance program and
compares this alternative with other agricultural policy alternatives in the context of the
13