Solving MPSGE Models Using GEMPACK

MPSGE is a non-algebraic language for the formulation of applied general equilibrium models. When you build a model using MPSGE you do not need to specify equations, but you instead work with a tabular representation of the model. The input tables make reference to source statistics, typically drawn from a single benchmark year. MPSGE input data describing your model characterize technology, preferences, tax rates, and factor endowments. These data, together with a set of equilibrium conditions, defines the equilibrium framework. MPSGE is a framework designed primarily to permit rapid prototyping of applied models. The compact, non-algebraic format can substantially improve productivity through the automated generation of demand and supply functions. This approach reduces the scope for errors and simplifies sensitivity analysis of results with respect to model structure. To take advantage of the MPSGE framework, a user must learn the syntax and conventions of the MPSGE language. The purpose of this paper is to introduce and document a new program (called MGE2GP) that can be used to convert MPSGE representations of models to GEMPACK. If you write down a model in MPSGE form, and create a data file for the model, you can use this program to convert your model to GEMPACK and then use GEMPACK software to solve the model. This paper contains detailed instructions for carrying out these steps. The paper contains a self-contained introduction to the MPSGE language, illustrated via several example models. The full version of the paper will contain detailed hands-on instructions for solving these example models using GEMPACK. The aim is to make the paper accessible to all general equilibrium modellers. In particular, we do not assume that readers are already familiar with MPSGE or GEMPACK. The example models can be solved using the Demonstration Version of GEMPACK which can be downloaded (at no cost) from the web.