Phillips trade-off from Phillips to Lucas: overview

Until the end of the 1960s, most Western economists believed in a stable trade-off between inflation and unemployment. The policy implication was that the government could choose a menu of inflation and unemployment rates along a given stable Phillips curve [see, for example, Samuelson and Solow (1960)].

This view was strongly shaken by Friedman (1968) in his presidential address to the American Economic Association. Friedman's basic challenge to a stable Phillips trade-off was theoretical: In an economy without money illusion, no significant real decisions should depend on the general level of prices. [Similar arguments were raised by Phelps (1967).] If all prices double, neither consumption, production, nor labor supply decisions should be affected. In particular, the rate of unemployment (a real variable) and the rate of inflation (the rate of change in a nominal variable) should not be systematically related.

Acceptance of this view raised the obvious question why empirically estimated Phillips curves displayed a systematic negative relationship between inflation and unemployment – at least until the end of the sixties. [See, for example, Phillips (1958) and Lipsey (1960) for the United Kingdom and Perry (1966) for the United States.] Friedman gave the following answer: When the rate of monetary growth accelerates and prices follow suit, employers realize that the price level is higher sooner than workers do. As a result, while nominally raising wages they can offer what amounts to a reduced real wage rate.