ABSTRACT

The economics profession has developed very little consensus regarding the nature of macroeconomic phenomena. One reason for this lack of consensus is the difficulty of passing from correlations in the historical data to cause and effect relations. Economists cannot perform the controlled experiments that render causation unambiguous, such as those used in the physical sciences. In macroeconomics, this difficulty in inferring structural relationships from reduced-form relationships appears most prominently in controversy over interpretation of the unemployment-inflation correlations summarized in Phillips curves.