The Method Used to Develop the Phillips Curve Is More Akin to That of the German Historical School

The Phillips curve is named after the British economist A. W. Phillips (1958), who in a pathbreaking article investigated the statistical relationship in the UK between the annual rate of change of money wages and the annual rate of unemployment. Later versions of the Phillips curve examined the relationship between unemployment or the rate of growth of output and, alternatively, the rate of change of product prices, and the deviation between actual and ‘expected’ inflation. 

Inasmuch as the Phillips curve is an important component of mainstream economics, its development remains a curious paradox. Logical positivism is the conventional methodology of the neoclassical mainstream and this methodological doctrine has been severely criticized by economists working within the Austrian methodological perspective. Logical positivism involves the construction of theory that is tested by empirical evidence. The resulting empirical evidence may lead to modification of the theory and further testing, but the initial theory construction always precedes empirical testing. The Phillips curve developed purely as an empirical relationship with only ad hoc theoretical rationalizations provided. Only later were attempts made to develop a theory that would ‘explain’ the statistical relationship. This method is more akin to that of the German historical school, of which criticism by the Austrian school has been much more severe. The series of currently recognized policy errors that followed from attempts to exploit the Phillips curve serves as an excellent but unfortunate example of the problems associated with ‘letting the facts speak for themselves.’

—Don Bellante, “The Phillips Curve,” in The Elgar Companion to Austrian Economics, ed. Peter J. Boettke (Aldershot, UK: Edward Elgar Publishing, 1994), 372.