Abstract
The GTAP final demand system has some known defects: the computation of the equivalent variation is not exact; with non-standard demand parameters, the equivalent variation may be grossly in error; the decomposition of the equivalent variation contains a nuisance term.
We find a further defect, that the upper-level demand equations are invalid. We revise the model to remove these defects.
Date of this Version
September 2000
COinS