The Evaluation of Econometric Modeling Languages: Syntax and Content

Econometric software can be evaluated on the basis of the facilities ostensibly offered, as is common in disciplinary software reviews. Alternatively, this software can be considered in terms of effectiveness: the degree to which it permits economists to perform research and other tasks. However, this approach immediately raises questions. Among these are: what are the tasks that should be permitted by the software? should the benchmarks be set simply to measure the accuracy of the results or should more general criteria be used? what types of benchmarks should be established? This paper addresses these questions.