Structure Overview Outdated

Structure overview #

MIRAGE is a model representing the world trade in a general equilibrium framework. Production and consumption in countries are precisely decomposed in regions thanks to input-output tables integrated in harmonised Social Accounting Matrices (SAM). Flows through regions and sectors as well as associated prices are represented explicitly and shocks on the equilibrated system allow to determine the deformation of the economy thanks to behavioural equations.

Data in MIRAGE #

MIRAGE relies mainly on GTAP data but also use other sources for important inputs such as protection levels, dynamics evolution, investments.

GTAP SAMs and trade data #

MIRAGE relies for SAMs on the GTAP database which currently provides data for 96 regions and 57 sectors. The model cannot run at this level of detail on a personal computer due to computation resource requirements and requires aggregation in a number of around 25 regions and 25 sectors.

GTAP also represents a strong network for research on global trade modelling. The GTAP official website constitutes an important source of information for CGE modellers.

Protection data #

Data on protection are particularly sensitive for in trade policy assessment. For simulations done with MIRAGE, we do not use GTAP aggregated tariff but generally compute our own tariff equivalents thanks to the database MAcMap-HS6 developed at CEPII and that represents bilateral applied tariffs for 5,113 products in the HS6 nomenclature.

See section: MAcMap database

Investment data #

A specific FDI data has been built at CEPII in order to represent transnational investment in the modelling.

See section: FDI database

Dynamics exogeneous data #

Dynamics is mainly driven by GDP projections from the World Bank and active population projections from ILOSTAT.

See section: Dynamics and Baseline (Outdated)

Behavioral equations #

Supply and demand #

The strength of the CGE framework is its foundation on microeconomic equations that represent the behaviour of agents in a consistent neo-classical framework. In MIRAGE, a representative agent is maximising its utility relatively to the perceived set of prices, and the demand in countries is complemented by investment demand under budget constraint and firms maximising their profit. The supply and demand of countries are entangled through international trade relations which make them interdependent.

See sections:

Production factors #

The representation of factor markets is also crucial to determine the mobility of production factors across countries and sectors, as well as their availability. There are in MIRAGE three type of factors common to all productive sectors: capital, skilled labour and unskilled labour. Additionally, primary sectors are also dependant on land (agriculture) and natural resources (for mining, forestry and fishing sectors).

See sections:

Integrated options #

Imperfect competition #

Firms behaviour are in the most simple framework considered as in perfect competition with non increasing returns to scale. MIRAGE can also be run in an imperfect competition framework representing multiple varieties and mark-up for firms.

See section Imperfect competition (Outdated)

Differentiation of quality #

Differentiation of quality is also an important distinctive feature that enable to introduce less substitutability between goods traded from developed countries and goods produced from developing countries.

See section Differentiation of quality

Foreign direct investments #

International capital mobility is also a feature that can be used to take into account foreign direct investment flows (FDIs) and their role in the economy.

See section Foreign direct investment

Other optional modules #

Several alternative versions of the model have been developed to take into account specific issues. They are not incorporated in the core version but can easily be implemented thanks to the explanations and the code made available by contributors on the following pages.

Model equations #

The model encompasses more than one hundred different equations, that are then replicated for each country and sector. The total number of equations during a modelling is generally around 1,000,000.

Model parameters #

The model relies on several sets of elasticities which are very important for results directions.