Almost thirty years ago Paul Samuelson and Bob Solow coined the term “Phillips curve” at the 1959 AEA meetings, reacting promptly to the publication of Phillips's (1958) article. For many years afterward Solow thought and wrote about the Phillips curve and many of the unsettled research puzzles that economists had struggled to resolve under that general heading. As Olivier Blanchard and Peter Diamond remind us, the Samuelson and Solow AEA paper (1960) was farseeing, anticipating many of the major issues that arose later when the Phillips curve started shifting. So it is fitting to take a look at the current state of the Phillips curve in economic research and its evolution since the seminal Phillips and Samuelson and Solow papers.

THE PHILLIPS CURVE NOW

To determine the difference between present views and those of the 1960s, and to highlight remaining puzzles, I take as my point of departure the current mainstream view of the U.S. inflation process. To find this mainstream view, you can look it up in any of the three best-selling intermediate macroeconomics textbooks. Here we find what I call the “triangle” model of inflation – inflation depends on three basic sets of factors: demand, supply, and inertia.

Formally, this model consists of two equations, the modern Phillips curve and a second equation, which, at least in my version, is a pure identity splitting the rate of nominal GNP growth in excess of potential output growth (this is “excess nominal GNP growth”) between inflation and changes in the output gap (i.e., in the log ratio of actual to potential output).