From Testiwiki
Jump to: navigation, search

The E3MG model is a multi-country econometric macro-sectoral simulation model for 20 world regions. The model combines the features of an annual medium- and long-term sectoral model estimated by formal econometric methods. E3MG solves for a very detailed set of inter-region and inter-industry relationships. The econometric grounding of E3MG enables the researcher to address the medium-term economic effects of E3 policies as well as the long-term effects of such policies.

An in-depth treatment of changes in the input-output structure of the economy over the forecast period incorporates the effects of technological change, relative price movements and changes in the composition of each industry's output. The model is designed for use in general global E3 analysis and for a wide range of policy analyses including: greenhouse gas mitigation, incentives for investment in renewable and other energy technologies, technological change, sustainable investment and sustainable household consumption.

The model provides annual E3 projections to the year 2020 and periodic projections (every 10 years) to 2100 for 20 world regions including Germany, France, Italy and the UK, EU15, EU10 and EU25, for industry output, investment prices, exports, imports, employment and intermediate demand at a 42-industry level including 20 service industries, and for consumer expenditures in 28 categories. The regional grouping includes 13 nation stares, OPEC and aggregates to the Annex I regional grouping.[1]


The advantages of the E3MG model lies in three areas:

  • Model disaggregation. The detailed nature of the model allows it to represent fairly complex scenarios, in particular scenarios which are differentiated according to sector and to country. Similarly, the impact of policies can be represented in a detailed way.
  • Econometric pedigree. The econometric grounding of the models gives it a better capability in representing and forecasting performance in the medium to long runs than CGE models. It therefore provides information which is closer to the time horizon of many policy makers than pure CGE models.
  • E3 linkages. An interaction (two-way feedback) between the economy, energy demand/supply and environmental emissions is an undoubted advantage over other models which may either ignore the interaction completely or only assume a one-way causation.

The E3MG 2.0 model has been developed by Cambridge Econometrics as a contribution in kind to the UK Tyndall Centre's CIAM modelling programme.[1]

Typical Model Applications:

  • Baseline analysis to 2100 with GDP, population, exchange rates, fiscal and monetary policies all exogenous.
  • Dynamic multiplier analysis, illustrating the response of the main economic indicators, industrial outputs and prices to standard changes in the assumption, e.g. changes in world oil prices, income taxes, government spending, and exchange rates.
  • Scenario analysis (differentiated according to sector and to country), across a range of greenhouse gas mitigation and energy policies at the global level, including carbon taxes and permit trading.
  • Analysis of long-term structural change in energy demand and supply and in the economy focused on the contribution of research and development, and associated technological innovation, on the dynamics of growth and change.[1]

Standard Model Specification:

Model structure:

E3MG comprises the accounting balances for products from input-output tables, for energy carriers from energy balances and for institutional incomes and expenditures from the national accounts. It includes environmental emission flows and 20 sets of time-series econometric equations. These can be grouped by:

  • aggregate energy demands (1 set of equations)
  • fuel substitution equations for coal, heavy oil, gas and electricity (4 set of equations)
  • commodity exports and imports (4 set of equations)
  • total consumers' expenditure (1 set of equations)
  • disaggregated consumers' expenditure (1 set of equations)
  • industrial fixed investment (1 set of equations)
  • industrial employment (1 set of equations)
  • industrial hours worked (1 set of equations)
  • labour participation (1 set of equations)
  • industrial prices (1 set of equations)
  • export and import prices (2 set of equations)
  • industrial wage rates (2 set of equations)
  • residual incomes (2 set of equations)
  • investment in dwellings (2 set of equations)

Each set of equation is estimated for the corresponding disaggregation level. For example, Employment is estimated in each region for each of the 41 E3MG 2.0 industries. This produces a rich set of econometrically estimated parameters: 558,647 parameters in a total of 17,015 equations. Each of the set of equations is estimated using panel-data techniques on time-series/cross-section data. Energy supplies and population stocks and flows are treated as exogenous.[1]

Sectoral coverage:

42 sectors:

(1) Agriculture etc, (2) Coal, (3) Oil & Gas etc, (4) Other Mining, (5) Food, Drink & Tobacco, (6) Textiles, Clothing & Leather, (7) Wood & Paper, (8) Printing & Publishing, (9) Manufactured Fuels, (10) Pharmaceuticals, (11) Chemicals nes, (12) Rubber & Plastics, (13) Non-Metallic Mineral Products, (14) Basic Metals, (15) Metal Goods, (16) Mechanical Engineering, (17) Electronics, (18) Electrical Engineering & Instruments, (19) Motor Vehicles, (20) Other Transport Equipment, (21) Manufacturing nes, (22) Electricity, (23) Gas Supply, (24) Water Supply, (25) Construction, (26) Distribution, (27) Retailing, (28) Hotels & Catering, (29) Land Transport etc, (30) Water Transport, (31) Air Transport, (32) Communications, (33) Banking & Finance, (34) Insurance, (35) Computing Services, (36) Professional Services, (37) Other Business Services, (38) Public Administration & Defence, (39) Education, (40) Health & Social Work, (41) Miscellaneous Services, (42) Unallocated.[1]

Consumption categories:

28 consumption goods

(1) Food, (2) Drink, (3) Tobacco, (4) Clothing and Footwear, (5) Gross Rent and Water, (6) Electricity, (7) Gas, (8) Liquid Fuels, (9) Other Fuels, (10) Furniture etc., (11) Household Textile etc., (12) Major Appliances, (13) Hardware, (14) Household Operation, (15) Domestic Services, (16) Medical Care etc., (17) Cars etc, (18) Petrol etc., (19) Rail Transport, (20) Buses and Coaches, (21) Air Transport, (22) Other Transport, (23) Communication, (24) Equipment etc., (25) Entertainment etc, (26) Exp. Rest and Hotel, (27) Misc. Goods and Services, (28) Unallocated.

Energy-environment module

E3MG considers interactions between the economy, energy demand/supply and environmental emissions. Energy and environmental industries and fuel types are highly disaggregated. The emissions captured are: CO2, SO2, NOx, CO, methane (CH4), black smoke, VOC, nuclear - air, lead - air, CFCs, N20, HFCs, PFCs, and SFCs.[1]

Dynamic structure:

  • Engle-Granger cointegration with error-correction dynamic models.
  • Annual forecasts to the year 2020.
  • Technical progress is measured by cumulative gross investment and data on R&D expenditure. Technology change in industry input-output coefficients through logistic growth.

Linkage between regions and countries:

For most commodities there is a European `pool' into which region supplies part of its production and from which each region satisfies part of its demands. The demand for a region's exports of a commodity depends on the relative prices, on an activity level of the main external EU export markets, and on the domestic demand for the commodity in all other EU regions, weighted by their economic distance.[1]

Market Structure:

Product market:

  • Imperfect monopolistic competition
  • Firms set prices (mark-ups over marginal costs)

Labour market:

  • Demand for employment partially adjusts to output growth, costs of labour and technology index.
  • Participation rate in the labour force depend on reservation wage.
  • Wage-setting decisions are depend on union activities across different regions of Europe. Unions choose wage rates to maximise utility subject to the labour-demand constraint.
  • Hours worked adjust ot technological change. Other adjustments to hours worked arise from short-run output adjustments.

Main model results:

Macro economic results (EU-wide and country level):

  • Industry output, investment prices, exports, imports, employment, intermediate demand
  • Consumer expenditure

Sectoral economic results (EU-wide and country level):

  • Industry output, investment prices, exports, imports, employment, intermediate demand (in 41 categories)
  • Consumer expenditure (in 28 categories)

Related environment results (macro/sectoral 20 world region level):

  • Energy carriers (12)
  • Energy users (21)
  • Atmospheric emissions (14 types, including 6 Kyoto GHGs)
  • Emission sources (10)
  • Energy technologies (25)[1]

Required technical infrastructure:

Construction and solution use the software package IDIOM, while the stochastic parameters are estimated using a general-to-specific (GETS) model estimation software designed in Ox.

Structure of Input Data and Data Sources:

  • Exogenous data: world data, population stocks and flows, economic policies (e.g. tax rates)
  • Parameter matrices, such as the parameters of the investment functions
  • Classification converters (sparse matrices)
  • Calibration values, lagged values, initialisation values
  • Data sources: OECD/IEA, OECD/STAN, World Bank, IMF, QUEST[1]

Links to other Models, Projects, Networks:

  • Part of UK Tyndall's CIAM
  • Links to models MDM and E3ME

Regional Scope:

E3MG20: World Regions

USA, Japan, Germany, UK, France, Italy, Rest EU-15, EU-10, Canada, Australia, OECD nes (not elsewhere specified), Russian Federation, Rest of Annex I, China, India, Mexico, Brazil, NICs[1]

See also


  1. 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 JRC: IA TOOLS. Supporting inpact assessment in the European Commission. [1]

Terry Barker, Jonathan Köhler, Haoran Pan, Rachel Warren, Sarah Winne (2004), Climate change mitigation policies and induced technological change: scenarios to 2100 using E3MG, to become a Tyndall Working Paper.