Greenhouse Gas Emissions Forecasting: Learning from International Best Practices

Discussion of Top-down, Bottom-up, and Hybrid Approaches to Energy-Economy Modelling

Energy-economy models generally range from detailed bottom-up models reflecting engineering economic details of a wide menu of technologies in each sector, to top-down models of the whole economy calibrated on historic data from a few to hundreds of sectors. Hybrid models -- those that combine the strengths of the bottom-up and top-down approaches -- are considered by many modelling authorities as optimal approaches to forecasting.^[40]

The essential element in a bottom-up model is that it is a model of individual units in a system, in which aggregate properties are then deduced from the behaviour of the individual units. For example, in a bottom-up model of soft drink consumption in a group of teenagers, we would look at the soft drink consumption of each teenager, and then add them up to get the group’s consumption. By contrast, a top-down model begins with a model of the aggregate; an attempt may be made to deduce properties of sub-units from the aggregate. In a top-down model of the group of teenagers, total soft drink consumption by the group is modelled, and then we try to allocate it among individuals.

Weaknesses in traditional top-down models include the fact that they do not explicitly represent the technologies in the energy system, so policies designed to influence technology evolution directly can only be crudely simulated at best. Bottom-up models are based on theoretical assumptions about human behaviour, with the result that their predictions do not represent the economic system. Because of their different structures and definitions of costs, the two types of models tend to predict very different economic structures. Top-down models usually predict high costs of GHG emission reduction policies while bottom-up models usually predict low costs. In order to address these weaknesses, modellers have attempted to create hybrid models that integrate the strengths of both the bottom-up and top-down approaches.^[41] Such hybrid energy-economy models have attempted to bridge the methodological schism between top-down and bottom-up approaches by meshing the description of the economy in terms of specific technologies (as in bottom-up models) into an integrated energy-economy model.

In Canada, both the E3MC and CIMS models are hybrid models used by a variety of actors. Hybrid models used in the U.S. and U.K. are NEMS and the U.K. MARKAL-Macro Energy Model.

40 Rivers, N. and M. Jaccard (2005). "Combining Top-Down and Bottom-up Approaches to Energy-Economy Modeling Using Discrete Choice Methods," The Energy Journal, Vol.26, No.1; Grubb, M (1993) "Policy Modeling for Climate Change: The Missing Models" Energy Policy, 21(3); Intergovernmental Panel on Climate Change (IPCC) (1996), IPCC Second Assessment Report: Climate Change 1995.

41 Grubb (1993), Hoffman and Jorgenson (1977), Jacobsen (1998), Bohringer (1998), and Koopmans and te Velde (2001) have all designed hybrid models.