GE model. In order to do so, the GE model is first used to simulate the 2010 baseline, using average
figures obtained from our econometric estimations.3 National supply curves are then used to calculate
country-level supply and production costs under the baseline and the reform scenarios. The fall in
domestic production caused by the reform is then introduced into the GE model which is used for
simulating the overall impact of the corresponding scenario. This iterative technique makes it possible
to include the information on supply in the different EU countries while taking advantage of the GE
model.
The model focuses on the EU. In practice, there are also large variations in production costs among
countries exporting sugar with the EU. That is, exports of third countries to the EU, including those
under preferential agreements, might be affected by the change in EU prices. Information on
production costs provided by Garside et al (2004), LMC (2004) show that some countries that export
to the EU, such as Guyana, St Kitts and Nevis, Barbados or Trinidad and Tobago are unlikely to
produce for the EU market after a large cut in EU intervention price. Their present production is
driven by the high internal EU price. However, since we do not consider the possible outcome of a
revision of the Cotonou regime, we assume that import quotas for ACP countries will remained
unchanged. We assume that a country benefiting from a duty free tariff quota will export sugar
(possibly purchased on the world market) to the EU market, as long as EU domestic price for raw
sugar is higher than the world price. In-quota imports with a positive tariff (i.e. imports under the
Traditional supply needs and Special preferential sugar provisions) will continue as long as the gap
between the EU domestic price for raw sugar and the world price exceed the tariff.
5. Policy changes simulations
The baseline. We first define a reference scenario or baseline which corresponds to the situation that
will take place in 2010, assuming the full implementation of the Agenda 2000, the June 2003 CAP
reform and the enlargement of the EU, i.e. without the 2005 reform of the sugar sector. The definition
of this baseline is of particular importance, since the June 2003 reform might have, in the absence of
other policy development, favored the production of C sugar in the most efficient regions. In the
3 In order to calibrate the sugar sector in our CGE model, we first determine the gross margin of both
beet productions (in and out-of-quota beets). We use input/output coefficients for a vector of
intermediate inputs, and returns to land from various sources, including Eurostat SPEL and the Farm
Accountancy Data Network. We assume that the sum of the margin of A&B and C beets is exhausted
in returns to the labor and capital bundle and quota rents. The econometric estimates of (7) are used to
calibrate the cross-subsidies between in-quota beets and C beets, assuming that the unitary implicit
subsidy on C beets adjusts to satisfy budget neutrality (i.e., the total implicit tax on in-quota beets
equals total implicit subsidy on C beets). This makes it possible to measure the value of the rent and
the value of the returns to the capital and labor aggregate. For the refining sector, a similar calibration
procedure is done, imposing that only A&B sugar beets are used to produce A&B sugar. Both A&B
and C sugar processing produces molasses and pulps. Unit labor costs cannot adjust in the processing
sector. Profit is exhausted in returns to capital and in quota rent. Again, econometric estimates of (7)
are used to calculate the rent.
15