subject to y1,t< y1t and y1,t+1< y1t + λy2,t , where λ represents the degree to which the producer estimates
that future quotas will be based on the present production of C sugar. Maximization in y2t leads to a
first order condition stating that Cmt = p2t +τ λ(p1,t+1-Cm(y t+1)). That is, the optimal production of C
sugar verifies the condition that Cmg=p2 plus a positive term. This term depends on future rents, i.e.
on future in-quota prices. While it is likely that producers expect future in-quota prices to be lower
than the present ones, the prospect that these prices will remain higher than the world price may
explain present production of C sugar.
Does this show that the C sugar is cross-subsidized? Formally, there is no direct linkage between the
in-quota sugar price and the supply of out-of-quota sugar, since the production of C sugar depends on
future in-quota prices, and not on present in-quota prices. However, if the "reference building"
behavior does not introduce a formal cross-subsidy, it may be one of the explanations for the relatively
high levels of C sugar produced during the recent years in the EU15, in spite of low world prices.
3. Econometric identification of production costs
Rents and production costs. The three cases presented above suggest that modeling of EU production
under the usual assumption, that producers maximize profit so that marginal revenue equals marginal
costs, may incorrectly represent EU supply response. If one calibrates the supply curve assuming such
a relation in a simulation model, and then uses the prices observed to infer marginal costs, this may
lead to construct an EU supply curve which lies below the actual one. The fall in EU production that
would take place under market liberalization could therefore be underestimated. Overall, this might
have a significant impact on the results obtained for world prices and trade.
Several authors have acknowledged that the different effects above should lead to a modification in
the traditional modeling of supply. Adenauer et al (2004) introduce shifts in the supply curve in order
to account for some of the phenomena described above. They identify the problem as being the
representative agent assumption. They account for the diversity of situations by using a larger number
of representative farms. Within their framework, the "insurance effect" described above affects
differently two producers with different levels of marginal costs. Once these individual behaviors are
aggregated, the resulting supply curve is such that beet growers behave like if their quota endowments
were higher than the actual ones.
The existence of a gap between marginal revenue and marginal costs for the aggregate producer is
central in our approach. A common feature of the three effects described above (fixed costs,
uncertainty on future yields and the asymmetry of gains/losses, expectation on future references) is
that the producer's behavior leads to conditions between the marginal cost Cm and the out-of-quota
price p2 of the type Cm=p2+θ, where θ is a positive function of p1, a negative function of p2 and a
positive function of the quota.
11