Wikis > Formulation Tools > Computable General Equilibrium Model (CGEM / CGE Model)
Primary Source: Suomalainen, K. (2006) Computable General Equilibrium Model SWOT Analysis in Report on the SWOT analysis of concepts, methods, and models potentially supporting LCA. Eds. Schepelmann, Ritthoff & Santman (Wuppertal Institute for Climate and Energy) & Jeswani and Azapagic (University of Manchester), pp 88-95

Level of analysis: Macro.

Assessed aspects of sustainability: Economic, possible to include environmental aspects.

Main purpose of the assessment: To assess macro-economical impact of policy changes.

Description of the methodology: Computable general equilibrium models are a class of economic models that use realistic economic data together with the abstract general equilibrium structure to simulate how an economy might react to changes in policy, technology or other external factors. In other words, CGEMs solve numerically the levels of supply, demand and price that support equilibrium across a specified set of markets. CGE models are a standard tool of empirical analysis, and are widely used to analyse the aggregate welfare and distributional impacts of policies whose effects may be transmitted through multiple markets, or contain menus of different tax, subsidy, quota or transfer instruments. Examples of their use may be found in areas as diverse as fiscal reform and development planning, international trade, and increasingly, environmental regulation (Wing).

Detailed description

Computer models are especially useful when analysing complex systems containing vast amounts of data. The benefits of computer models include explicitness, infallible computing and comprehensiveness (Sterman, 1991). In equilibrium models demand elasticity is introduced. The user specifies a demand curve rather than a single value for the demand. Based on the analysis, the demand level can change in response to changes in prices. The three conditions of market clearance, zero profit and income balance are employed by CGE models to solve simultaneously for the set of prices and the allocation of goods and factors that support general equilibrium. Demand drives the production processes and since production varies among the different markets, this in turn influences the economy’s growth. However, most CGE models conform only loosely to the theoretical general equilibrium paradigm. For example, they may allow for:

–          non-market clearing, especially for labour (unemployment) or for commodities (inventories),

–          imperfect competition (e.g. monopoly pricing),

–          demands not influenced by price (e.g. government demands),

–          a range of taxes,

–          externalities, such as pollution.

CGEMs can be either static, quasi-dynamic or dynamic. A static model generates a solution of the problem for a single point in time, neglecting any temporal development. Quasi-dynamic and dynamic models analyse the system over longer periods of time (multiple time periods). In quasi-dynamic modelling the system is optimised for one period at a time, with the solution for one period forming the starting point for the solution of the next period. In fully dynamic models the solution for all periods is calculated simultaneously and an optimum for the entire time horizon is found. In environmental analysis static models have been more common due to the fact that the additional information given by a dynamic model is not necessarily relevant from the environmental point of view. Also for environmental analysis it is important to have a relatively disaggregated model, i.e. have many economic sectors. Dynamic macro models tend to have the economy as the only sector, simply because it easily becomes technically difficult to solve with several sectors.

A static CGEM gives an image of the situation in the start and end year of the analysis. The initial “shock effects” of an introduced tax, policy or change in price on international markets leading to turbulence due to changed relative prices until a new equilibrium has been found, are not shown. A dynamic CGEM, on the other hand, describes the development from one state of equilibrium to another. CGEM in general are strongly linked to national economic micro theory and often use the structure and data of national accounts e.g. input-output tables. Price elasticities are also central parameters. They are sector specific and should be derived from empirical data and econometric methods.

CGEMs are based on several simplifying assumptions such as perfect competition, perfect information, no external effects and no common goods – the models are built on strictly theoretical formulation. These are standard assumptions in the basic theory national economic analysis. The system boundaries generally coincide with those of the data sources, i.e. the national (or EU) level of economical politics under study.


CGEMs are most often used for policy simulations, where a certain economic development is assumed for the studied period and the model simulates the impact on economy for different policy measures (Dervis et al. 1985, Spadaro 2007). One can also do consequential analysis of a certain policy measure and see what the impact would be given various economic development scenarios. In relation to LCA, more specifically consequential LCA, CGEMs have recently been used for identifying marginal effects of the studied changes, providing data for the inventory assessment, e.g. the marginal long-term effects of changes in crop demand on global land-use (Kloverpris et at. 2007). Other examples of applications include trade policy analysis (Francois and Shiells, 1994), social cost analysis (Hazilla and Kopp, 1990) and capturing the impacts of specific markets on surroundings (Kehoe and Kehoe 1994, Gilbert and Wahl 2002). Studies on energy and emission permits are increasingly common applications of CGEMs(Bhattacharyya 1996), two examples of which are given below.

Example: GEM-E3

At EU-level the model GEM-E3 (General Equilibrium Model for studying Economy-Energy- Environment interactions) was developed as a multinational collaboration and currently represents 21 World regions (World model) or 15 European countries (Europe model), linked through endogenous bilateral trade and environmental flows. The European model is being extended to include associated countries and Switzerland.

The model computes simultaneously the competitive market equilibrium and determines the optimum balance for energy demand and/or supply and emission/abatement. It has been developed to include several imperfections (real world approximation) such as imperfect competition of the labour market and certain sectors. Applications of the model have been carried out for several Directorate Generals of the European Commission and for national authorities.

The GEM-E3 model includes all simultaneously interrelated markets and represents the system at the appropriate level with respect to geography, the sub-system (energy, environment, economy) and the dynamic mechanisms of agent’s behaviour. It formulates separately the supply or demand behaviour of the economic agents which are all considered to optimize their own objective while market derived prices guarantee global equilibrium. The model considers explicitly the market clearing mechanism and the related price formation in the energy, environment and economy markets. The prices are computed as a result of supply and demand interactions, while different market mechanisms, in addition to perfect competition, are allowed.

The most important results provided by GEM-E3 are (EC, 2002): dynamic annual projections in volume, value and deflators of national accounts for each country; full input-output tables by country; distribution of income and transfers in the form of a social accounting matrix by country/region; employment, capital and investment by country or region or sector; atmospheric emissions, pollution abatement capital, purchase of pollution permits and damages; consumption matrix by product and investment matrix by ownership branch, public finance, tax incidence and revenues by country or region; full trade matrix for the considered countries/regions.

Example MacGEM

MacGEM is a global marginal abatement cost simulation model. It consists of a set of marginal abatement cost functions for carbon emissions originating from fossil fuel use. The model aims at evaluating compliance costs and permit trading equilibriums for the first commitment period of the UN Framework Convention on Climate Change UNFCCC. Emission trading equilibriums are computed by seeking a price for which total market excess permit supply is zero. Excess supply of each of the 15 world regions/countries in the model depends on its marginal abatement cost function and assigned amount of emissions. The marginal abatement cost functions are estimated on data generated with the GEM-E3-World general equilibrium model. MacGEM also allows for the introduction of trading restrictions like for instance limited accessibility of the Kyoto flexible mechanisms like Joint Implementation (JI) and Clean Development Mechanism (CDM).

In MacGEM the GDP in 2010 of country i is defined as: GDPi=GDPiBAU−Ci(Ri) where GDPiBAU denotes the projected Business-As-Usual GDP level for 2010 and Ci(Ri) denotes theemission abatement cost for country i for reducing its emissions with Ri tons compared toprojected BAU emissions. Actual emissions in 2010 are defined as 2010 BAU emissions minusabatement: Ei =EiBAU –Ri. The emission abatement cost function denotes the GDP lossincurred by country i if it has to cut back its carbon emissions with Ri tons by 2010. Theselosses include, among others, the costs of fuel switching, the cost of investing in more efficienttechnologies and insulation costs to increase fuel efficiency in private houses and buildings.

The cost function is assumed to be twice continuously differentiable; strictly increasing (Ci>0for Ri>0) and strictly convex in abatement (Ci′′>0). Hence marginal abatement costs are risingas more emissions are abated. Furthermore, it is assumed that the first unit of abatement is free(Ci(0) = 0 and Ci(0) = 0) and that it is infinitely costly to abate the last unit of emissions(limRi→Ei(BAU)Ci′(Ri)=+∞).

A market for carbon emission permits is created by assigning emission targets (Assigned Amount Units AAUi) to every region and allowing them to trade emission reductions. The possibility of permit trading affects a country’s GDP in the following way:


GDPi = GDPiBAU – Ci(Ri) + p[ AAUi – Ei ] = GDPiBAU – Ci(Ri) + p[AAUi – EiBAU + Ri]


Every country can choose between reducing its emissions more than required by the quotum AAUi and selling the surplus in the permit market at unit price p, or reducing its emissions less than required and buying additional permits in the international market. Assuming price taking behaviour and ignoring constraints on the trading volumes, a free trade market equilibrium for permit trading is defined as a vector of emission reduction efforts such that every individual country maximises its expected GDP in 2010. The first-order necessary and sufficient condition for this maximisation problem says that every country should reduce its carbon emissions up to the point where its marginal abatement cost is exactly equal to the market price: Ci′(Ri) = p. These first-order conditions define well-behaved, continuous and increasing emission reduction supply curves: ρi(p) =Ci-1(p) since Ci′ is strictly monotone, continuous and strictly increasing in abatement. Excess supply for permits is defined as follows: XSi(p) AAUi−Ei=AAUi−EiBAUi(p). If XSi(p) <0, the actual emissions of region i in 2010 are higher than the Assigned Amount Units and it has to import emission permits in order to comply with its emission reduction commitment. If XSi(p) >0, country i is exporting emission permits since its emissions are lower that its AAU. A permit market equilibrium for the set of countries S is defined as a price level p*≥0 for which total excess supply is non-negative: ΣjЄSXSi(p*)≥0. The excess supply framework can easily be extended to account for transaction costs and limited accessibility for e.g. CDM and JI projects, by altering the production supply function as follows: ρi(p) =αCi′-1 ([1]p), where α denotes the accessibility (for example 30%) and β the proportional transaction cost (for example 20%) that is incurred when implementing a bilateral JI or CDM project.


Scope of assessment (what is being assessed)

CGEMs are most relevant for assessing complex, data-intensive systems, such as the global problems they have typically been used for. They are especially strong in assessing activities related to markets and trade. In such a situation all relevant aspects can be taken into account, including externalities, as long as they can be linked to the flows defining supply or demand.

In relation to LCA, the detail of environmental analysis typically is limited to the accounting of one or few specific indicators, such as quantifying GHG emissions or changes in land use. However, an inventory may, in principle, be constructed (by broadening the scope) to include several, if not all, relevant aspects from an environmental point of view, and the results linked to environmental impact assessment methods, presenting the possibility of linking CGEM results with LCA.

Methodology (robustness, validity & reliability)

In its “computablility”, the method has three main strengths:

–          Assumptions made by the modeller can be stated in written documentation and be subject to review. The used data can be assessed by professionals.

–          Computers makes no logical errors, but infallibly compute the logical (quantitative) consequences of the given assumptions.

–          A computable model can be all-inclusive and able to interrelate a great number of facts simultaneously.

Thus, CGE models are robust methodologically based on well-developed theory (neoclassical microeconomics) and implementation. When applied to well-known global problems, as in the examples described above that are based on public data from established institutions, the results have high validity and reliability, taking into account the underlying assumptions.


Scope of assessment

CGEMs give a pathway to reach a certain goal under certain assumptions and constraints. However one must be careful in forecasting and predicting actual behaviour, which a CGEM does not give.

Due to its macro-level scope, many details are left out, which is true from an LCA viewpoint as well. Although a full inventory for environmental impact assessment could in principle be constructed, it would still be limited by the sectoral aggregation of CGEMs, and is generally not sufficiently detailed for directly providing input to typical LCA micro-level studies.


CGEMs are limited by assumptions of neoclassical economics: market clearance, perfect competition, entirely price driven behaviour of agents and perfect foresight, attributes that do not characterize reality. Also the considered time slices (often annual for example) do not capture certain relevant effects that vary seasonally or hourly (e.g renewable energy production). In addition data for calibration can be hard to get – the base year may not serve for calibration purposes if it was an extreme year, e.g. extremely rainy or hot. A base year does not show trends either, which must be assumed exogenously. Data collection and modelling can be cost-intensive and time-consuming tasks.

Opportunities for broadening and deepening LCA

CGEMs already play an important role in European energy-environment-economy research and present the opportunity to study wider aspects, include several variables to these interconnections. These could portray socio-economic aspects for example. In EU research CGEMs are beneficial for long-term strategic planning for sustainable development.

The Environmental Technology Action Plan (ETAP) offers an opportunity for the use of a CGEM for making long-term scenarios with the different technologies that are to be introduced and the policies that are to support them.

Threats for broadening and deepening LCA

A threat to the use of CGEMs could be its apparent “black box” nature and the necessity of professionals for the use of the tool. Results may be perceived as not being transparent enough and thus not trustworthy. It might also be desirable to optimize other variables e.g. resource use, which could be a threat for CGEMs and promote other types of optimization approaches.

Literature/Internet links

Bhattacharyya, S. C., 1996, Applied general equilibrium models for energy studies: a survey, Energy Economics 18, pp. 145-164.

Dervis, K., Melo, J. De, Robinson, S., 1985, General equilibrium models for development policy, A World Bank research publication, Cambridge University Press, Cambridge, MA, USA.