Assessing Modelling Approaches to the Distributional Effects of Macroeconomic Policy

This article outlines a set of desirable properties of an adequate model for assessing distributional impacts of macroeconomic policy packages, and uses these as criteria to evaluate four model types commonly used for such assessments, namely, fixed ratio, econometric, CGE and microsimulation models. The desirable properties fall into five broad groups: being able to represent policy levels used in the policy packages, allowing for both price- and quantity-clearing markets, integrating financial stocks, representing short-term dynamics, and providing measures of confidence in a model's output. Of the model types reviewed, CGE models perform the best on most criteria, in spite of their substantial flaws of non-verifiability and lack of short-term analysis.