Overview of CGE modelling

CGE models are commonly used tools for policy analysis. CGE models are particularly helpful when considering how changes in one part of the economy flow through to other industries. They can be used to estimate the economywide and distributional impacts of almost any policy or industry change, including regulatory change, taxes or subsidies, environmental policies, tariffs, preference shifts, etc.

Such models typically consist of:

1. A database that represents an economy in a certain year based on input-output (IO) tables. The database specifies the interactions and relationships between various economic agents including firms, workers, households, the government and overseas markets.
2. Behavioural parameters governing agents’ responses to relative price changes (e.g. elasticities).
3. A system of equations that define the model specification or theory, which is generally based on standard economic assumptions, but not necessarily constrained by them (for example, in the always-and-everywhere attainment of equilibrium after shocks are imposed).

From an initial equilibrium where demand equals supply in all factor, final demand and intermediate input markets, the system is then ‘shocked’ by changing one or more variables that represent a policy change or other change in economic conditions.

By comparing the pre- and post-shock databases, we can then observe the effects of the shock in question in terms of changes to GDP, employment, wages, industry output, etc. Static CGE models consider only ‘before’ and ‘after’ the policy shock. There is no ability to consider the nature of the adjustment path between equilibria.

A dynamic CGE model – such as MDG6NZ – allows the user to examine in each intervening period (usually each year) how variables adjust from the time when a shock is implemented to the time when all its effects have worked through the economy (which may be several years).

In order to capture adjustment lags, MDG6NZ incorporates sticky wages and delayed investment responses to capital prices, rather than assuming instantaneous factor market adjustments.