Environmental Engineering Reference
Variables (axes F1 et F2 : 32,14 %)
F1 (17,73 %)
Fig. 2.2 Principal component analysis performed on the Auvergne farms set (axes 1 and 2)
Qualitative Assessment of Jointness
More sophisticated procedures are required to proceed with steps two and three of
our framework (qualitative and quantitative assessment of jointness), so we turned
to SEAMLESS-IF. SEAMLESS-IF does not directly assess the jointness in the
production system between the supply of commodity and non-commodity outputs.
The FSSIM-MP component represents the farmer's behaviour with regard to the
possible relations between the production factors inside the production process of
commodity outputs. Environmental non-commodity outputs are modelled as exter-
nalities (their supply is non-volunteer) or as constraints in the production process
of commodity outputs (see Chapter 5 for more details).
The first simulations for some farms in four regions already suggest interesting
insights about jointness at the farm level despite the lack of operational indicators.
Figure 2.3 illustrates the modelled nitrate (NO 3 −1 ) leaching with regard to farms
income (per hectare) for some SEAMLESS-IF crop farms in Auvergne. When the
sample of farms contains more than a few individuals, the relationship between
both variables is not that simple.
Figure 2.4 depicts the marginal effect of environmental subsidies on farm income for
FADN farms in the Auvergne region, for 2000 and 2004. The estimated coefficient is
similar to coefficient b 5 in Table 2.3 , but estimated for each individual farm (in Table 2. 3
it is estimated on the whole sample of FADN farms). It is obvious from the Figure that
this marginal effect is not the same for the 2 years, but we can notice an increased effect
for some farms and a decreased effect for others. Aggregating these individual