Some results from the Australian version of TERM
TERM developments
Documentation of TERM
TERM for other countries
Download Australian TERM model and data
TERM (The Enormous Regional Model) is a "bottom-up" CGE model of Australia which treats each region as a separate economy. The key feature of TERM, in comparison to predecessors such as MMRF, is its ability to handle a greater number of regions or sectors. The first TERM master database distinguished 144 sectors and 57 regions (nearly corresponding to the Australian Statistical Divisions). More recently, TERM has been extended to represent 172 sectors in 206 statistical sub-divisions (SSDs): this allows us to split major cities into regions, and represent water catchment and tourism regions. The high degree of regional detail makes TERM a useful tool for examining the regional impacts of shocks (especially supply-side shocks) that may be region-specific. Finally, TERM has a particularly detailed treatment of transport costs and is naturally suited to simulating the effects of improving particular road or rail links.
Even though TERM is computationally efficient, it would be slow to solve if a full 144-sector, 57- or 206-region database were used. In practice, we must aggregate sectors or regions so that [the number of sectors] plus [the number of sectors] does not exceed 100. The TERM database programs facilitate this aggregation. The choice of sectors or regions to aggregate is application-specific. For example, we could design aggregated regions which followed the boundaries of climatic zones or watersheds, or which highlighted the distinction between metropolitan and rural regions. Similarly the sectoral aggregation would be tailored to a particular simulation.
Since September 2006 we have been developing a new TERM feature: a "top-down" disaggregation of employment and income within each Statistical Division to the SLA (Statistical Local Area) level. That is, results originally generated for the 57 SDs may be broken down according to Australia's 1379 SLAs. Although we have reservations about the realism of results at SLA level, we anticipate that the facility may prove useful when used with care. For example, it would enable us to say how residents of Sydney's western suburbs (a group of SLAs) fared relative to other Sydneysiders. See COPS archive item TPMH0074 for an example of this approach.
Subsequently, the SLA data gathered from the 2006 census became the foundation for the SSD region representation mentioned above. In practice, we could prepare a full bottom-up database for any combination of the 1400+ SLA regions in the 2006 census. However, the SSD representation for most practical purposes has been a sufficient representation of small regions.
A group at the Australian Productivity Commission used a version of TERM to see if water trading could reduce the economic impact of water shortage. This was published as Peterson, Dwyer, Appels & Fry(2005), Water Trade in the Southern Murray-Darling Basin, The Economic Record 81. An earlier version of the paper is here.
The original version of TERM is a comparative static model. It shows, for a single year, the differences produced in the regional economies by changes in taxes, techonology, tariffs and other exogenous variables. Glyn Wittwer has recently developed a multiperiod version of TERM. This incorporates dynamic features similar to the MONASH model:
For an application of dynamic TERM, see: Dynamic general equilibrium analysis of improved weed management in Australia's winter cropping systems, by Glyn Wittwer, David T. Vere, Randall E. Jones, Garry R. Griffith, The Australian Journal of Agricultural and Resource Economics, December 2005 - Vol. 49, Issue 4, pp 363-377, Blackwell.
Enquiries about consultancies using the Australian or Chinese versions of TERM should be addressed to Dr. Glyn Wittwer. He is currently using dynamic TERM for projects with Plant Health Australia and Geoscience Australia. To apply TERM to other countries, contact Mark Horridge.
See COPS archive item TPMH0057 for a technical
description of the model and database. An earlier version of the same paper is working paper no. G-141. Other applications are described in
COPS archive items TPGW0050 and TPGW0076. A published article is:
Mark Horridge, John Madden and Glyn Wittwer (2005), The impact of the 2002-2003
drought on Australia, Journal of Policy Modeling, vol. 27, issue 3, pages 285-308
A complete worked example of TERM database production can be found here.
See a comparison of CoPS regional CGE models here.
Some results from the Australian version of TERM
Versions of TERM have been prepared for Brazil, Finland, China, Indonesia and Japan. The Chinese version is described on this page. A simulation using the Indonesian version is described in COPS archive item TPDP0038. See also TPGW0076. Some Brazilian TERM results are presented in COPS archive item PPJB0078.
You can download and run a miniature version of the original Australian TERM yourself if you have source code GEMPACK. Download this ZIP file and study the README.TXT it contains.