© Scott McDonald & Karen Thierfelder
CGEMOD has four families of models; all are coded using the General Algebraic Modeling System (GAMS), which is our preferred programming language. All the models in these families use the same notation, many of the same behavioural relationships, the same methods for handling data (inputs and outputs) and the same approach to the organisation of the code.
There are two single-country CGE model families: the STAGE (Static Applied General Equilibrium) family, which is a suite of full feature CGE models, and SMOD (Simple MODel), which is suite of training models that have features that were ‘state of the art’ some 20 years ago.
There are two global CGE model families: the GLOBE family, which is a suite of full feature global CGE models that can be calibrated using the WIOD or GTAP databases, and R23, which is a global variant of the classic 123 model.
The opens source variants of the STAGE and GLOBE families are setup as ‘template’ models. While they are full feature models they are intended as the starting point for customisation of the behavioural relationships by the user.
The standard version of the STAGE and GLOBE models are setup to analyse trade, fiscal and general economic policy issues. Variants of the STAGE and GLOBE families include those for analysing energy, migration, income distribution, labour market, agricultural and other policies. In addition to the comparative static versions they are also recursive dynamic versions.
The online training CGEMOD offers uses SMOD, STAGE 1 and GLOBE 1 (see training page). Although these models are used during training, the training programmes are designed with the objective that ‘graduates’ will make changes to the models going forward.
CGEMOD also offers customised advanced training to uses of models in the STAGE and GLOBE families of models (see advance training)