Abstract

We interpret ceteris paribus conditions as the conditions necessary to conducting an experiment. "Ceteris paribus" is thus not a hold-all for whatever we do not know, but a listing of the various decisions taken in moving from a theoretical hypothesis to a testable proposition. The decisions range from modeling in a certain way to selecting a particular functional form or estimation technique. They embody best knowledge/best practice. Debate about the meaning and importance of any test result must center on these decisions; hence they should be laid bare. We give a detailed example from recent macroeconomics, Lucas' test of the natural rate hypothesis