Ceteris Paribus: Some Notes on Methodology

At the heart of any analytical process lies simplification or abstraction, the whole purpose of which is that of making problems scientifically manageable. In the economic system we recognize, of course, that depends on everything else, and also that is always changing. For the requirements of analytical rigor to be met fully, we should work only with dynamic Walrasian systems. But real problems can rarely be studied in dynamic terms and almost never in true Walrasian terms. Real problems require the construction of models, and the skill of the scientist is reflected in the predictive or explanatory value of the model chosen. We simplify reality to construct these models, but the fundamental truth of interdependence must never be forgotten. The interdependence of the economy imposes certain restrictions on any analyst, and each model stands or falls on the test of its consistency with the general equilibrium world. This paper will demonstrate that many highly sophisticated models fail this test, and for the same reason. Specifically, I shall try to show that there exists a common error in the use of the Marshallian device of ceteris paribus.