CGE Baseline
Welcome to the CGE Baseline model comparison project.
Since the early 1990s, dynamic Computable General Equilibrium (CGE) models have increasingly been used to analyse impacts of shocks to the system to economic activity in the medium- and long-term future. Quantitative policy analysis is typically presented in relation to a scenario – i.e. a numerical projection – that assumes the economy will change under the influence of future changes in the socioeconomic drivers (and policy assumptions that are beyond the domain of interest of the policy analysis). These so-called baseline scenarios provide a reference point for evaluating the impact of policy changes or other events. Constructing better baselines serves two purposes. First, it can improve the insights into plausible and internally consistent projections for future economic activity in absence of policy changes or other shocks. Such projections can be extended to related indicators that are connected to economic activity, such as energy security, poverty, nutrition and climate change damages. Secondly, it can help explain the differences between insights obtained by global dynamic CGE models insofar these are grounded in the underlying baseline characteristics.
This page documents the progress on an informal community-led model comparison project to advance the state of baseline projections in CGE models.
This comparison exercise started with a workshop organised by Organisation for Economic Co-operation and Development (OECD) and the Center for Global Trade Analysis (GTAP) in January 2018. Since then, different author teams have carried out in-depth analysis of the baseline assumptions in the various models, the results of which are presented in a special issue in the Journal of Global Economic Analysis (JGEA) that was published in June 2020.
The first phase of the project has led to a number of recommendations for follow-up (see the overview paper by Dellink et al in the JGEA special issue for more details):
- It would be highly valuable to follow up the knowledge created here with a multi-model quantitative intercomparison exercise. Participating modelling teams would identify and run a set of scenarios with partially harmonised assumptions.
- It would be useful to better document existing CGE modelling baseline scenarios and the key assumptions underlying the numerical projections. We intend to create an on-line space for modelling teams to document a “model wiki” that contains the main model features, and link that to an online inventory of recent baseline projections on the current website.
- It would be extremely useful to improve the categorization of baseline scenarios. Once the key information for different existing baseline scenarios is available, broad categories of scenarios can be constructed. This activity involves finding suitable clusters of scenarios and creating narratives that support the cluster.
- Finally, there could be room for user-led activities. Some priority areas identified in the first phase include
- Identifying how to improve the way CGE models differentiate between short-term and long-term responses to shocks to the economic system.
- Identifying how to better integrate fiscal and financial aspects of the economy.
- There is a constant need to improve model assumptions; better econometric estimations of key model parameters, such as elasticities of substitution and transformation, and the parameters that drive dynamic behaviour over time, incl. income elasticities. The quantitative model comparison exercise suggested above should be able to identify the main research priorities in this respect. Better input-output tables for developing countries, and especially African economies, is also a key priority here.
CALL FOR INPUT:
Researchers that produce global dynamic economic projections using CGE models are invited to contact CGEBaselines@purdue.edu to share the key information on their model and the baselines they produce. Emails from interested researchers should include the following information: (1) model name, (2) the institution that owns the model, (3) a contact person for the model, (4) the model’s website (if applicable), (5) a published paper describing the model, and (6) a paper or website describing the latest baseline. If you want to play an active role in the next phase, please indicate which activity you’d like to be involved in and what role you’d like to play.
Further information on this project will be made available on this webpage shortly.
Last Modified: Wednesday, August 16, 2023