ZUR WISSENSCHAFTSTHEORIE DER ÖKONOMETRIE

SUMMARY This contribution intends to discuss the question whether econometrics indeed conform to the methodological demands, outlined by P opper in his Logic of Scientific Discovery for empirical research. Part I deals shortly with the goals and methods of econometrics. It is then argued that the falsification criterion can be used on stochastic hypotheses in a moderated version only, not in the stringent form as postulated by Popper (Part II). This necessitates, however, that tautologies, contrary to common practise, be ruled out. Using the Cobb‐Douglas‐function as an example it is shown that it is possible in econometrics to exempt an originally non‐circular concept from testing by introducing a residue variable and so rendering in effect a tautology (Part III). Finally, the question whether aggregate data of macro‐economic surveys meet the validity criterion is discussed and negated (Part IV).