The MyGTAP Project

The MyGTAP project is a user based initiative. Researchers and Center staff are contributors to the project, however, the project is an independent, user motivated initiative, housed outside of Purdue University’s GTAP Center.

The MyGTAP initiative was founded in 2013 to extend the standard GTAP model to include improved representation of government and household incomes and expenditures.  MyGTAP was initially funded by a grant from the Bank of the Netherlands Partnership Program (BNPP) and World Bank to enhance modeling the vulnerabilities of African countries to external shocks.  Continuing interest in further extending the MyGTAP model motivated the development of the MyGTAP website, which is supported by the original authors of the MyGTAP model and data tools, Terrie Walmsley and Pete Minor.

The Motivation for MyGTAP

The Standard GTAP model (Hertel 1997) created a strong baseline for comparing the global impacts of trade policy for twenty years.  At the heart of the GTAP philosophy is the regional household, which blurs the links between household incomes, taxes, and expenditures.  This allowed researchers to abstract from the distributional and budgetary consequences of policies – a wise choice given the data collection challenges. The Standard GTAP model, therefore, depended on symmetric data compilation, that is, it seeks the lowest common denominator in terms of data availability – data for one, must be available for all.  As a result, many researcher have been forced to alternative, single country models, to capture the rich insights provided by explicitly linking household’s expenditures and incomes and government taxes\revenues.  However, this approach is not without its drawbacks, the use of single country models is often at the cost of losing potentially important global feedback mechanism.

What is the MyGTAP Project?

The MyGTAP project is an effort to bridge the gap between the Standard GTAP model and the advantages and insights gained by utilizing a single country modeling approach.  The “My” in MyGTAP is a recognition of the departure from the global GTAP database; the user will have to bring their own data to the analysis. However, at the heart of our approach is the GTAP database – all trade, production and macro data must be reconciled with the GTAP data. The MyGTAP modeling approach leverages advanced coding methods to allow users the flexibility to integrate individual country data into the database and model in an effort to enhance modeling insights, while maintaining the standard GTAP data structure for other countries or regions.

Support and Training

Current technical support for the model is limited to technical questions related to the coding and running of the data tools and model.  Understanding the economics and data are fully the user’s responsibility.  Users familiar with the GTAP model, either through GTAP training programs or other training will find the MyGTAP model familiar.  Users interested in training would do well to attend a course offered by the GTAP Center on CGE modeling and the Standard GTAP model or look into hosting a customized courses offered by ImpactECON. Documentation for the model is lengthy and an effort has been made to make it as accessible as possible.

Contributions and Appreciation

MyGTAP, like most CGE models, was built on contributions of many individuals and organizations.  The authors are grateful for the important contributions including the base model code of the GTAP model developed by the Global Trade Analysis Project (GTAP) and SplitCom code created by Mark Horridge of Monash University, employed frequently in the data splitting programs in MyGTAP.  The authors are grateful to Mark Horridge for his continued support which has include valuable advice for integrating MyGTAP into RunGTAP and coding suggestion and innovations to support the unique, flexible, structure of MyGTAP.  The MyGTAP model also benefited from discussions with Scott McDonald and Karen Thierfelder on the GLOBE model.  The MyGTAP model, while unique in many ways, shares attributes with the GLOBE model.  In fact, users interested in employing the MyGTAP framework within GAMS will find many common features and overlap.  A GAMS program to translate the MyGTAP data into GLOBE is currently being developed by MyGTAP and GLOBE developers.  Early users of the model and data program provided useful feedback including: Khalid Siddiq, Mary Burfisher, Angel Aguiar, Godfrey Mahofa, Christopher Onyango and Anna Strutt.  Will Martin of the World Bank has provided invaluable support by sponsoring the original grant establishing MyGTAP under the BNPP GTAP Africa Project.