Abstract

Economics like other sciences needs to solve "the problem of induction" as a methodological problem to become an experimental science. The founders of the paradigm of econometrics were to resolve the problem by the theories of "probability and statistical inference". The present paper tried to show that the paradigm of econometrics failed to achieve this aim. Not compatible with what the founders of the paradigm of econometrics thought which this problem can only be solved by technical reasons based on probability and statistics, the present paper tried to show that this failure has arisen from a methodological reason; not solving the problem of induction, which itself originated from an epistemological problem; the problem of causation. To support this claim, it attempted to demonstrate that the paradigm of econometrics, in spite of its varying techniques, was concentrated on the Humean conception of causality which failed to solve the problem of induction. The solution for changing methodology was laid in epistemological considerations, not only in varying statistical techniques. Most criticisms concerning econometrics have not paid enough attention to this controversial point and focused on statistical techniques. This failure will challenge the claim of the paradigm of econometrics on the scientific policy – making in economics.