CGE modelling offers a flexible and robust framework to quantify and explain how the economy evolves in response to changes in the economic environment. Changes in population size and characteristics, investment spending and capital stocks, government spending and taxation policies and changes in the foreign sector are just some of the “policy shocks” that can be investigated. CGE modelling is a globally popular analytical tool, recognised as the “gold standard” for informing policy formulation and analysis through scenario simulation.
CGE stands for computable general equilibrium – the most basic form of a CGE model is built on neoclassical microfoundations and calibrated using data for the economic area(s) of interest (e.g. national accounts data for a single country model). The primary advantage of CGE analysis over more traditional and simpler multiplier (input-output) analysis is that CGE models impose supply constraints and therefore relative prices are endogenous and explicitly solved for.
Despite the rich detail and robust underlying theoretical structure of CGE models, their predictive power should not be overstated. Data for many parameters often does not exist and the model is only calibrated (as opposed to estimated) using the data that is available. Thus, while CGE models produce quantitative results, their magnitudes should not be relied upon to the extent that econometrically estimated results can be (i.e. to a certain level of statistical significance).
CGE models are most useful when taking a kind of “differences-in-differences” approach (using the term more loosely than in the usual statistical technique sense) For instance, one might compare the welfare effects on heterogenous households in a region when there is an export boom of a product that is intensively produced in that region under alternative scenarios such as when in-migration of persons in relevant occupational groups (i.e. with relevant skill sets) is allowed or not allowed. The welfare effects across household types could be quite different in each scenario and would indicate the “winners and losers” of the export boom if in-migration was likely (or unregulated). Such conclusions would be more reliable and useful than just quoting quantitative model predictions of increases in regional GDP etc. arising from the export boom.